GUIDELINES FOR THE DIAGNOSIS OF LATENT TUBERCULOSIS INFECTIONglobaltb.njms.rutgers.edu/downloads/products/guideltbi.pdf · guidelines for the diagnosis of latent tuberculosis infection

GUIDELINES FOR THE DIAGNOSIS OFLATENT TUBERCULOSIS INFECTION in the 21st Century


PO Box 1709 • 225 Warren Street • Newark, NJ 07101-1709

CME Certified Monograph2nd Edition

GUIDELINES FOR THE DIAGNOSIS OF LATENT

TUBERCULOSIS INFECTION IN THE 21st CENTURY

CONTINUING EDUCATION CERTIFIED MONOGRAPH

Sponsored by the University of Medicine & Dentistry of New Jersey (UMDNJ), and

the New Jersey Medical School Global Tuberculosis Institute

Original Release Date: June 2002 • Activity Update: April 1, 2008 • Expiration Date: March 31, 2010

Nursing credit for this activity will be provided through March 31, 2010

This activity is supported by educational grants from Monarch Pharmaceuticals for the first edition release and JHP

Pharmaceuticals for the release of the second edition.

TARGET AUDIENCE

This activity is designed for internists, pediatricians, pulmonologists, infectious disease specialists, public health and

preventive medicine specialists, nurses, and other health personnel interested or involved in tuberculosis diagnosis,

treatment and prevention of tuberculosis and latent tuberculosis infection.

LEARNING OBJECTIVES

Upon the completion of this activity, participants should be able to:

� Describe the role of tuberculin testing in low prevalence countries

� Review how tuberculins are developed, manufactured and validated

� Recognize minor disparities in commercially available tuberculins and the necessity of serial testing with the same antigen

� Explain the protocol for administering and reading tuberculin skin tests

� Correctly interpret repeated tuberculin skin tests

� Examine the role of tuberculin reactions produced by cross reactions with non-tuberculous mycobacteria

� Differentiate the use of interferon-γ release assays when compared to tuberculin skin testing

� Discuss the role of the nurse in the diagnosis of latent TB infection

METHOD OF INSTRUCTION

Participants should read the learning objectives and the activity in its entirety. After reviewing the material, complete

the post-test consisting of a series of multiple-choice questions.

Upon completing this activity as designed and achieving a passing score 70% or more on the post-test, participants

will receive a continuing education credit letter and test answer key four weeks after receipt of the registration and

evaluation materials.

Estimated time to complete this activity as designed is 3.5 hours.

c1

ACCREDITATION

Physicians: UMDNJ–Center for Continuing and Outreach Education is accredited by the Accreditation Council for

Continuing Medical Education to provide continuing education for physicians.

UMDNJ–Center for Continuing and Outreach Education designates this educational activity for a maximum of 3.5 AMA PRA

Category 1 Credit(s)™. Physicians should only claim credit commensurate with the extent of their participation in the activity.

Nurses: UMDNJ–Center for Continuing and Outreach Education is an approved provider of continuing nursing

education by NJSNA, an accredited approver by the American Nurses Credentialing Center’s Commission on Accreditation.

This activity is awarded 3.5 contact hours (60 minute CH).

Provider approved by the California Board of Registered Nursing, Provider Number CEP 13780.

This activity was peer-reviewed for relevance, accuracy of content and balance of presentation by Lee B. Reichman, MD,

MPH and Rajita Bhavaraju, MPH; and pilot-tested for time required for participation by Anju Budhwani, MD, Henry S.

Fraimow, MD, DJ McCabe, RN, MSN, and Lillian Pirog, RN, PNP.

DISCLOSURE

In accordance with the disclosure policies of UMDNJ and to conform with ACCME and FDA guidelines, individuals in a position

to control the content of this education activity are required to disclose to the activity participants: 1) the existence of any financial

interest or other relationships with propriety entities producing health care goods and services, with the exemption of non-profit or

government organizations and non-health care related companies, within the past 12 months; and 2) the identification of a

commercial product/device that is unlabeled for use or an investigational use of a product/device not yet approved.

The faculty and editors listed below have declared that they have no significant financial relationships or affiliations

to disclosure:

John-Manuel Andriote, MS Kitty Lambregts, MD, PhD, MPHRajita Bhavaraju, MPH Alfred Lardizabal, MDGeorge Comstock, MD, DrPH Richard Menzies, MD, MScKaren Galanowsky, RN, MPH Sheldon L. Morris, PhD

Elsa Villarino, MD, MPH

Lee B. Reichman, MD, MPH is a consultant for and shareholder of Cellestis, Inc.

The pilot-testers listed below have declared that they have no significant financial relationships or affiliations to disclosure:

Anju Budhwani, MD DJ McCabe, RN, MSNHenry S. Fraimow, MD Lillian Pirog, RN, PNP

OFF-LABEL USAGE DISCLOSURE

This publication contains information about the test, T-SPOT.TB, which has not been approved by the U.S. Food and

Drug Administration for the detection of M. tuberculosis as of the date of this publication.

DISCLAIMER

The views expressed in this publication are those of the faculty. It should not be inferred or assumed that they are

expressing the views of Monarch and JHP Pharmaceuticals, any other manufacturer of pharmaceuticals, or UMDNJ.

The drug selection and dosage information provided in this publication are believed to be accurate. However, the reader

is urged to consult the full prescribing information on any drug mentioned in this publication for recommended dosage,

indications, contraindications, warnings, precautions, and adverse effects before prescribing any medication. This is

particularly important when a drug is new or infrequently prescribed.

Copyright © 2008 UMDNJ-Center for Continuing and Outreach Education. All rights reserved including translation

into other languages. No part of this publication may be reproduced or transmitted in any form or by any means, electronic

or mechanical, including photocopying, recording, or any information storage and retrieval systems, without permission in

writing from UMDNJ-Center for Continuing and Outreach Education.

To review UMDNJ’s privacy policy: http://ccoe.umdnj.edu/general/privacypolicy.html.

Please direct continuing education related questions to UMDNJ at 800-227-4852 or email [email protected].––––––––––––––––––––––

This activity is accessible via Internet Explorer and Netscape browsers. A high speed connection is highly recommended.

You must have Adobe Acrobat Reader version 4 or above installed on your computer to view PDF files. If you do not have

Acrobat Reader, you can download it for free at Adobe.com.

Page composition and typography by Williams Production Group, Annandale, Virginia.


Continuing Education Certified Monograph2nd Edition

Edited by Lee B. Reichman, MD, MPH

John-Manuel Andriote, MS (First Edition)Rajita Bhavaraju, MPH (Second Edition)

IntroductionLee B. Reichman, M.D., M.P.H.

Relevance of the Tuberculin Test in Low-Prevalence CountriesKitty Lambregts, M.D., Ph.D, M.P.H.

Developing a Clinical Tuberculin Test: Tuberculin Characteristics, Reactivity and Potency

Sheldon L. Morris, Ph.D.

Randomized Clinical Trials of Specificities of Commercially Available Tuberculins

Elsa Villarino, M.D., M.P.H.

Administering and Reading Tuberculin Skin Tests;Interpreting Repeated Tuberculin Skin Tests

Richard Menzies, M.D., M.Sc.

Tuberculin Sensitivity Produced by MycobacteriaOther than the Mycobacterium Tuberculosis Complex

George Comstock, M.D., Dr.P.H.

Interferon-γ Release Assay for Detectionof Tuberculosis Infection

Alfred Lardizabal, M.D.

The Role of the Nurse inDiagnosing Latent TB InfectionKaren Galanowsky, R.N., M.P.H.

GUIDELINES FOR THE DIAGNOSIS OF LATENT

TUBERCULOSIS INFECTION IN THE 21st CENTURY

A special symposium was held in Miami Beach, Florida. It was

chaired by Lee B. Reichman, M.D., M.P.H., and the faculty featured

six distinguished authorities on the diagnosis of latent tuberculosis

infection.

This monograph presents suggested guidelines that emerged

from the meeting, which are intended for the practicing clinician,

and emphasize the administration and interpretation of

tuberculin tests as part of the diagnosis and treatment of latent

tuberculosis infection.

The symposium and monograph was supported by an

unrestricted educational grant from Monarch Pharmaceuticals.

Due to unprecedented interest, authors were asked to revise

and update chapters for a second edition, which was supported

by an unrestricted educational grant from JHP Pharmaceuticals.

John-Manuel Andriote was the technical editor of the original

monograph. Rajita Bhavaraju revised the 2nd edition.

SUGGESTED CITATION

To Cite the Entire Monograph

Reichman LB, Bhavaraju R, eds.Guidelines for the Diagnosis of Latent Tuberculosis Infection in

the 21st Century, 2nd Edition. Newark: New Jersey Medical School Global Tuberculosis

Institute, 2008.

To Cite a Specific Chapter

Menzies R. Interpreting Repeated Tuberculin Skin Tests. pages 38-46. in Reichman LB,

Bhavaraju R, eds: Guidelines for the Diagnosis of Latent Tuberculosis Infection in the 21st Century,

2nd Edition. Newark: New Jersey Medical School Global Tuberculosis Institute, 2008.

Guidelines for the Diagnosis of Latent Tuberculosis Infection in the 21st Century, 2nd Edition

George Comstock, M.D., Dr.P.H. passed away on July 15, 2007 after a long illness. We were particularly gratified that

even though quite ill, Dr. Comstock saw fit to bring his chapter, “Tuberculin Sensitivity Produced by Mycobacteria other

than Mycobacterium tuberculosis Complex,” up to date for the second edition of this monograph.

Much of our knowledge about latent tuberculosis infection is directly due to Dr. Comstock’s work. His research and

application of its results, has served as the basis of the diagnosis and management of this condition.

Dr. Comstock was a good friend and mentor to us all. It is with gratitude, appreciation, and respect that we dedicate this

monograph to his memory.

—Lee B. Reichman, M.D., M.P.H.

April 1, 2008

Dedication to George Comstock, M.D., Dr.P.H.

Guidelines for the Diagnosis of Latent Tuberculosis Infection in the 21st Century, 2nd Edition

Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Lee B. Reichman, M.D., M.P.H.

Relevance of the Tuberculin Test in Low-Prevalence Countries . . . . . . . . . . . . . . . 9

Kitty Lambregts, M.D., Ph.D., M.P.H.

Developing a Clinical Tuberculin Test: Tuberculin Characteristics,

Reactivity and Potency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18


Randomized Clinical Trials of Specificities of Commercially

Available Tuberculins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

Elsa Villarino, M.D., M.P.H.

Administering and Reading Tuberculin Skin Tests . . . . . . . . . . . . . . . . . . . . . . . . 32

Dick Menzies, M.D., M.Sc.

Interpreting Repeated Tuberculin Skin Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38


Tuberculin Sensitivity Produced by Mycobacteria Other than the

Mycobacterium Tuberculosis Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

George Comstock, M.D., Dr.P.H.

Interferon-γ Release Assay for Detection of Tuberculosis Infection . . . . . . . . . . . 57

Alfred Lardizabal, M.D.

The Role of the Nurse in Diagnosing Latent TB Infection . . . . . . . . . . . . . . . . . . . 62

Karen Galanowsky R.N., M.P.H.

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

Activity Self-Assessment Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Registration Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

Activity Evaluation Form . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

CONTENTS

Guidelines for the Diagnosis of Latent Tuberculosis Infection in the 21st Century, 2nd Edition 1

For more than three decades, treatment of persons

with latent Mycobacterium tuberculosis infection to

prevent active disease has been an essential component of

TB control in the United States. The symposium on which

this monograph is based was organized to address the

confusion that has been generated by periodic outbreaks of

TB in the United States. Looking at these outbreaks, one

thing is clear: They were unnecessary and likely could

have been prevented if those at high risk were targeted,

tested and treated.

EVOLUTION OF TREATMENT GUIDELINES

In 1965, treatment for latent TB infection was first

recommended for those with previously untreated TB,

tuberculin skin test (TST) converters and all children

under age three with a positive tuberculin test result.

The recommendations were broadened in 1967 to

include all who were TST positive (>10mm) and their

close contacts.

Treatment guidelines were developed in 1974

regarding pre-treatment screening and monitoring to

minimize the risk for hepatitis and to exclude low-risk

persons over age 35 as candidates for treatment. The

guidelines were further revised in 1983 to recommend

routine clinical and laboratory monitoring for persons

older than 35 or with increased risk for hepatoxicity.

In the years since HIV/AIDS was first reported in 1981,

TB has reemerged as a tremendous threat to individuals

with compromised immune systems. Many of those

infected with HIV are co-infected with TB. For this reason,

two months of RIF and PZA (2RZ) were recommended in

1998 for co-infected HIV-positive persons, and

subsequently for those who were not HIV-infected.

Treatment guidelines for latent TB—what used to be

called “chemoprophylaxis”—have continued to be

refined. In 2000, nine months of Isoniazid was decreed

to be more effective that six months, and 2RZ were

deemed equal to nine months of Isoniazid. But in 2001

2RZ was de-emphasized, due to its liver toxicity, in favor

of nine months of Isoniazid. In 2004, rifampin alone for

4 months was also recommended.

TARGETED TUBERCULIN TESTING AND

TREATMENT OF LATENT TB INFECTION

As the rate of active tuberculosis in the United States

has decreased, identification and treatment of persons

with latent infection who are at high risk for active TB

have become essential components of the nation’s TB

elimination strategy.

In its 2000 report, Ending Neglect: The Elimination of

Tuberculosis in the United States, the Institute of Medicine

(IOM) recommended precisely what this symposium is

intended to address: an increased emphasis on targeted

tuberculin testing and treatment of latent TB infection.

“To begin advancing toward the elimination of

tuberculosis,” said the IOM report, “aggressive new

efforts must be implemented to identify those who are at

greatest risk of disease through targeted programs of

tuberculin skin testing coupled with treatment for latent

tuberculosis infection.” It continued, “The question now

confronting the United States is whether another cycle of

neglect will be allowed to begin or whether, instead,

decisive action will be taken.”

THE SYMPOSIUM AND GUIDELINES

The 2002 symposium, held in Miami Beach, Florida,

brought together clinicians and researchers with

considerable expertise on the diagnosis and treatment of

latent TB infection. Each of them prepared and

presented a paper that describes various aspects of

diagnosing and treating latent TB infection. These papers

are collected in this monograph, and have been updated

by the authors for this second edition.

Because the FDA approved an effective gamma

interferon release assay, we have added a chapter on

QuantiFERON®-TB Gold In-Tube.

EXECUTIVE SUMMARY

2 Guidelines for the Diagnosis of Latent Tuberculosis Infection in the 21st Century, 2nd Edition

RELEVANCE OF THE TUBERCULIN TEST IN

LOW-PREVALENCE COUNTRIES

The incidence of TB disease is declining in the United

States and other industrialized countries, and in fact the

U.S. has reached the elimination phase of the TB

epidemic. An important challenge, however, lies in the

fact that while the number of U.S.-born Americans

infected with TB continues to shrink, the pool of infected

foreign-born individuals grows larger. Among the

problems associated with imported TB are drug-resistant

bacteria, microepidemics among sub-populations

(including illegal immigrants, the homeless and

prisoners) and a reduction in the overall effectiveness of

treatment for latent TB infection.

The positive effect of adequate TB control—i.e.,

decreasing incidence—hampers TB control in the

elimination phase because professional expertise on TB

fades away. This results in diagnostic delay, improper

diagnostic tools, inadequate treatment, poor infection

control and inadequate guidance for patients. Other

challenges to diagnosing and treating latent TB infection

include the lack of cultural understanding among many

immigrants of treating a disease without symptoms. The

overlap of the TB and HIV/AIDS epidemics present

another major challenge.

Targeted tuberculin testing for latent TB infection is a

strategic component of TB control in low-prevalence

countries in that it identifies persons who are at high risk

for developing TB and would benefit from treatment of

latent TB infection. In particular, tuberculin skin testing

(TST) is useful for:

� Contact tracing

� Providing pre-exposure baseline TST results for

exposure groups, such as health care workers

� Screening of persons with clinical conditions

associated with progression to active disease

� Screening of risk groups for TB

� Screening travelers to high-prevalence countries

� Skin-testing of symptomatic patients, and

� Measuring annual risk of infection and program

impact

QuantiFERON®-TB Gold In-Tube has similar specific

advantages.

Despite its usefulness, the TST also has consequences

and limitations. In particular:

� A TST result, whether negative or positive, may

add valuable information in symptomatic patients

� A test result just below the cutoff point for a

specific target group should provide guidance to

both patient and health care providers by raising

the level of suspicion so they will consider TB if

there are symptoms

� In some instances—such as with liver disease or

among the elderly who may long have been

infected—the physician may decide not to start

treatment for latent TB infection but rather to

offer close monitoring and proper instruction to

the patient

� There may be false positive or negative reactions

due to either technical or biological causes

� Boosting of tuberculin sensitivity

DEVELOPING A CLINICAL TUBERCULIN TEST:TUBERCULIN CHARACTERISTICS, REACTIVITY

AND POTENCY

Tuberculins are complex mixtures of culture filtrate

components derived from sterilized cultures of tubercle

bacilli. The predominant form of tuberculin used in the

United States is Tuberculin PPD, a protein precipitate of

M. tuberculosis culture filtrate. Tuberculins are widely

used to detect exposure to M. tuberculosis because they

induce delayed-type hypersensitivity immune reactions

in individuals who have been sensitized to

mycobacterial antigens.

There are four primary stages in the clinical

development of new tuberculin preparations:

(1) Preclinical development and testing

(2) Investigational New Drug (IND) stage

(3) Biologics License Application stage

(4) Post-licensure stage

During the critical IND stage, the potency of the new

product is evaluated by comparing its skin test reactivity

to the skin test responses induced by a standard dose of

PPD-S, the U.S. Standard PPD. When the dose of the

new product that is bioequivalent to PPD-S has been

determined, the specificity and sensitivity of the new

tuberculin is often assessed by testing it in at least three

populations: persons known to be infected with M.

tuberculosis, persons living in areas with low

mycobacterial infection and persons living in areas with

high atypical mycobacterial infection rates.


If the potency, specificity and sensitivity of the new

product are shown to be appropriate in these clinical trials,

these data become the foundation of a Biologics License

application. To ensure consistent reactivity after licensure,

the potency of each commercial tuberculin lot must be

shown to be bioequivalent to PPD-S prior to distribution.

RANDOMIZED CLINICAL TRIALS OF

SPECIFICITIES OF COMMERCIALLY AVAILABLE

TUBERCULINS

The tuberculin skin test (TST) is the standard method

for diagnosing infection with M. tuberculosis. The test

involves intracutaneous injection of 5 tuberculin units

(TU) of purified protein derivative (PPD) by the

Mantoux technique. Two companies manufacture

PPD tuberculin in the United States: JHP Pharmaceuticals

(Aplisol®) and Pasteur Mérieux Connaught (Tubersol®).

Despite FDA regulations for production and

standardization of PPD tuberculin, there have been

concerns that these commercial PPD products may vary

in performance. Clusters of unexpected positive

reactions or suspected false-positive results involving

both products have been reported in the medical

literature, to the FDA and to the Centers for Disease

Control and Prevention (CDC).

The accurate diagnosis of infection is important to

ensure that infected persons receive appropriate

evaluation and treatment, and that uninfected persons

are not exposed to unnecessary evaluation and treatment.

The possibility that one or both of the commercial PPD

products may have an unacceptably high rate of false-

positive reactions prompted a study. The study

compared the specificity—i.e., the percentage of

uninfected persons correctly categorized—and the

distribution of reaction sizes of the two commercial PPD

reagents among a population of subjects who, because of

their history, were at low risk for infection with M.

tuberculosis. Besides skin testing with the two

commercial PPD reagents, subjects were skin tested with

PPD-S, the “gold standard” PPD test.

The randomized, double-blinded trial among a total

of 1,596 volunteers revealed that:

� Testing with Tubersol® produced smaller

reactions, and with Aplisol®, larger reactions, than

PPD-S, but these differences did not affect TST

interpretations

� The specificities of Aplisol® and Tubersol® were

equally high and similar to that of PPD-S

� Both Aplisol® and Tubersol® correctly classified

comparable numbers of persons not infected

with TB

� Either commercial product may be used with

confidence for TST

These study findings suggest:

� It is important to remember that the biological

variability in response to TST, as well as technical

differences in admininistering and reading the test

will result in increases or decreases of <6 mm in

95% of subjects

� It is also important to remember that erythema and

bruising at the TST reaction site are not indicative

of positive reactions and should be disregarded

when interpreting the TST

� The sensitivity and specificity of the TST for the

detection of latent TB infection (LTBI) is unknown

because no test can provide formal proof of the

presence or absence of LTBI

� Even when done in the most reliable way possible,

the TST remains an imperfect diagnostic tool and

should not replace clinical judgment

� Tubersol® demonstrates the greatest number of

non-reactors and thus would appear to be the least

suited for a screening test. Most importantly,

routine TST in a low-TB incidence area is of limited

use because the majority of positive results will be

false-positive

� With respect to rates of false positive results and

based on the results of this study, there are

probably no differences between the different

tuberculins

� At present, the most accurate method to diagnose

LTBI is the Mantoux TST, which requires:

– Testing a targeted high-risk population

– Properly administering a dose of a standardized

tuberculin preparation

– A trained professional correctly interpreting any

observed reaction

� An increase in the percentage of positive skin test

results is possible after a change of tuberculin

preparations


Results of the study show that the key to approach this

issue is to think probabilistically by:

� First, using the available information to estimate

the likelihood of disease

� Then assessing the potential benefits and risk of

the proposed interventions (x-rays, sputum test,

drug therapy)

� Recognizing that sometimes the best strategy is to

wait for more information via repeat testing

A new laboratory based blood test with good

specificity and sensitivity, QuantiFERON®-TB Gold In-

Tube, has recently been FDA approved and will likely

find its use in screening of low risk reactors as well as

contacts for whom Mantoux tuberculin skin testing is

indicated.

ADMINISTERING AND READING TUBERCULIN

SKIN TESTS

Injecting tuberculin material intradermally into a

person previously infected with M. tuberculosis will result

in infiltration of previously sensitized lymphocytes from

circulating peripheral blood. At the site of the injection,

CD4 and CD8 T-lymphocytes, monocytes and

macrophages will accumulate. These release

inflammatory mediators, which produce edema and

erythema. Although this results in increased blood flow,

the locally increased metabolic activity of these

inflammatory cells results in relative hypoxia and

acidosis, which may be severe enough to lead to

ulceration and necrosis.

Certain guiding principles can help in appropriately

administering and reading tuberculin skin tests:

� Indications

– Test only those in whom therapy for latent TB

infection is indicated—namely, the “high-risk”

reactors:

• Recent infections (contact, conversion)

• Increased reactivation risk (HIV, diabetes,

abnormal chest x-ray, etc.)

� Contraindications

– Known documented severe reaction to TST in

the past

– Documented prior positive test result (TST

should be done if there is a history of

undocumented prior positive TST result)

� Administration

– Use only the Mantoux method, i.e., intradermal

injection on forearm

– Use 5-TU of PPD—bio-equivalent to PPD-S

– Anergy testing is not useful nor recommended

� Reading

– All reading must be done by trained health

professionals

– All readings must be made at 48-72 hours

– Measure the transverse diameter of induration

– Record result in millimeters

� Adverse Reactions

– Severe adverse reactions are rare

– Local allergic reactions occur in 2-3%, and are

not related to the actual tuberculin result

– Strongly positive reactions with blistering:

• Should be covered with a dry dressing

(to prevent scratching)

• Cold compresses may be soothing

• Corticosteroids (topical cream) are not effective

INTERPRETING REPEATED TUBERCULIN

SKIN TESTS

The use of repeated tuberculin tests to detect new TB

infections in high-risk populations has often resulted in

problems of interpretation. This is because tuberculin

reactions may change size because of random variation of

the test, or because of a real biologic increase. But it also

may be due to boosting or conversion.

� Random (Non-Specific or Chance) Variation—

When multiple tuberculin tests are administered

and read, resulting test-to-test differences will

include differences due to administration and

reading as well as inherent biological variability.

– Can account for changes of 1 to 5 millimeters in

size (bigger or smaller)

– So 6 millimeters is criterion to distinguish true

biologic changes in reaction

� Boosting (Two-step testing)—This phenomenon

is defined as an increase in tuberculin skin reactions

of at least 6 mm following repeat tuberculin testing

and unrelated to new mycobacterial infection. It is

believed to occur when cell-mediated response has

waned, resulting in an initially negative tuberculin

reaction, but the tuberculin test stimulates


anamnestic immune recall. Boosting is often seen

among elderly persons.

– Seen if two tuberculin tests repeated in absence

of new infection

– Maximum if interval between two TSTs is

one week

• Less if only two to three days between TST

• Still detected after one year or more

– Associated with old age and foreign birth

(remote TB infection)

• Also with BCG vaccination (common in

foreign born)

• And with non-tuberculous mycobacteria

(common in southern USA and foreign-born)

– Non-specific reaction—risk of true infection

is lower

• Risk of disease is also lower

– Management—medical evaluation and chest

x-ray

• Benefit of therapy, and, therefore, need for

therapy is less

� Conversion—Tuberculin conversion is defined as an

increase in tuberculin reactions of at least 6mm

following repeat tuberculin testing and is due to new

mycobacterial infection.

– Defined as an increase of TST following new

mycobacterial infection

• Could be true TB infection, or BCG

vaccination, or nontuberculous mycobacteria

– Definition based on TST size

• Increase of 6 mm—most sensitive but less

specific

• Increase of 10 mm—less sensitive but more

specific - generally used

• Increase of 15 mm—much less

sensitive but more specific—not generally

used

– Prognosis—high risk of disease

• Highest risk if conversion following known

TB contact

– Management—medical evaluation and chest

x-ray

• Therapy for LTBI strongly recommended for

all ages

TUBERCULIN SENSITIVITY PRODUCED BY

MYCOBACTERIA OTHER THAN THE

MYCOBACTERIUM TUBERCULOSIS COMPLEX

“Nontuberculous mycobacteria” is a name suggested

for this numerous and diverse group of mycobacteria. As

the name suggests, the group is defined by exclusion, i.e.,

mycobacteria other than M. tuberculosis. Although

nontuberculous mycobacteria were recognized as early as

1885, they were rarely considered in the first half of the

20th century. By century’s end, however, they had

become well known because of the widespread use of

diagnostic cultures of M. tuberculosis and because of the

disease they caused among persons whose immune

systems had been compromised by HIV.

Studies of the similar reactions caused by both TB and

non-TB mycobacteria sparked renewed interest in the

ability of nontuberculous mycobacteria to provoke

sensitivity on the tests used to detect TB infection much

like TB itself. The outcome was a clearer understanding

that not all tuberculin test reactions are caused by

infection with M. tuberculosis. In short:

� A positive tuberculin test is not always due to

M. tuberculosis: During the decades after World

War II, it was demonstrated that reactions to the

tuberculin tests were not always due to infections

with M. tuberculosis.

� Role of nontuberculous mycobacteria:

Infections with a variety of nontuberculous

mycobacteria are common, especially in the

warmer parts of the world. They cause reactions to

the tuberculin test that tend to be smaller than

those due to M. tuberculosis.

� Probability, not certainty: There is no way to

differentiate all individuals who are infected with

M. tuberculosis from those infected with other

mycobacteria (including M. bovis BCG). The size of

the tuberculin test reaction indicates the

probability that it was caused by infection with

M. tuberculosis.

� Accurate measurements of reactions are

essential: To find the optimal cut-point between

positive and negative reactions, or to apply the

CDC/ATS recommended cut-points with maximal

effectiveness, requires accurate measurements of

induration. Specifically, this means there should be

a smooth distribution of reaction sizes without

undue proportions of reaction sizes ending in 5 or 0.


� What can be seen, can be measured: Margins

of induration should be visualized by examining

the arm in proper lighting or by marking the

margins by the ballpoint pen method. In either

case, it is highly desirable to use a gauge or calipers

that do not allow the scale to be seen until after the

measurement has been made.

� Selecting a locally optimal cut-point: The

mirror-image method of estimating the proportion

of reactions caused by M. tuberculosis indicates

the optimal cut-point between positive and

negative reactions, taking into consideration the

local relative frequency of tuberculous and

nontuberculous infections.

INTERFERON-γ RELEASE ASSAY FOR

DETECTION OF TUBERCULOSIS INFECTION

Until recently, the standard and only method for

immunologic diagnosis of M. tuberculosis infection has been

limited to the tuberculin skin test (TST). However, because

purified protein derivative of tuberculin contains many

antigens that are shared with other mycobacteria, the skin

test does not reliably distinguish LTBI from prior

immunization with Mycobacterium bovis bacilli Calmette-

Guérin (BCG) or infection with environmental

mycobacteria. False-negative results in the setting of host

immunosuppression has limited also its utility. In addition,

cutaneous sensitivity to tuberculin develops from 2 to 10

weeks after infection and the TST requires two encounters

with a health care professional which often causes logistical

problems if not inconvenience. Finally, skilled personnel are

essential for proper placement and interpretation of the test.

Countering many of the concerns associated with the

TST, the QuantiFERON®-TB test (QFT) was approved by

the US Food and Drug Administration (FDA) in 2001 and

the and current generation of this test, QuantiFERON®-

TB Gold In-Tube (QFT-G), received final approval from

the FDA in 2007. Like the TST, QFT measures a

component of cell-mediated immune reactivity (CMI) to

purified protein derivative (PPD) from M. tuberculosis as

well as M. avium intercellure. However, as a blood assay,

the QFT requires a single patient visit, and because it is an

ex vivo test, it does not boost anamnestic immune

responses. The interpretation of the whole-blood

interferon gamma release assay (IGRA) is less subjective

than the TST, and the test is less affected by prior BCG

vaccination and reactivity to non-tuberculous

mycobacteria than the TST.

The principles of QuantiFERON®-TB Gold In-Tube

Test are incubation of whole blood with antigens,

measurement of IFN-γ by ELISA, and interpretation of

test results.

THE ROLE OF THE NURSE IN DIAGNOSING

LATENT TB INFECTION

The nurse plays a vital role in the diagnosis of latent

TB infection by:

� Providing tuberculin skin testing for high-risk

individuals within a community, and

� Ensuring that those individuals who have a positive

tuberculin skin test are medically evaluated and

treated for latent TB infection to completion.

By doing these two things, future cases of TB will

be prevented.

Targeted tuberculin skin testing and treatment of

those individuals with latent TB infection is a public

health activity, which has significant health, social and

economic benefits. To accomplish this, the nurse must

integrate the core functions of public health into

interventions and strategies at the individual, community

and health care system levels. This requires knowledge

and competencies regarding:

� The nursing process

� The diagnosis of latent TB infection and disease

� Tuberculin skin testing administration, reading,

and interpretation

� Community assessment

� Adherence

� Epidemiology

� Regulations and legal mandates

� Patient education that is culturally and

linguistically appropriate

� Collaboration

� Networking

� Evaluation

Confronted with the challenges at the individual,

community and health care system level, nurses utilize

knowledge and competencies to intervene appropriately

and prevail to achieve significant outcomes.


Lee B. Reichman, M.D.,

M.P.H.

Dr. Reichman is Professor of

Medicine, Preventive Medicine

and Community Health, and

Executive Director of the New

Jersey Medical School Global

Tuberculosis Institute in Newark,

New Jersey

For more than three decades, treatment of persons

with latent Mycobacterium tuberculosis infection to

prevent active disease has been an essential component of

TB control in the United States. This symposium was

organized to address the confusion that has been generated

by periodic outbreaks of TB in the United States. Looking

at these outbreaks, one thing is clear: they were

unnecessary and likely could have been prevented if those

at high risk were targeted, tested and treated.

CONFUSION ABOUT TB TESTING

There has been confusion about how to target

tuberculin testing and treat latent TB infection, evidenced

in written and e-mail correspondence received by the

New Jersey Medical School Global Tuberculosis Institute:

� “I have taken two tests and they both come out

puffy and red and the clinic said they are positive. I

just enrolled my daughter in preschool and now I’m

not allowed to help in the school. What is TB and

how come my tests are coming up positive?”

� “I am a physician and have a 20-year-old patient

recently arrived from Finland. As is the practice

there, she received BCG as a child approximately

18 months ago. A recent PPD resulted in an 18

mm reaction. Her chest x-ray is normal and she

has neither a history of TB exposure nor immuno-

suppressive conditions. She has, however, traveled

to Russia in the past few years on at least two

occasions. I am aware of the general

recommendations about BCG and PPD

interpretation but the Finish government has

provided this lady with written statements that her

PPD is a result of the BCG and that no treatment is

advised. What do you think? Are there

organizational recommendations (e.g., WHO) that

I can refer to in this matter?”

� “My sixteen-year-old daughter has tested positive

for TB. I am very much concerned. She was asked

to take Isoniazid 300 mg tablets for the next six

months. Is it really necessary to go through this?

She has no symptoms. According to our physician

she tested positive because of some immunization

shot given in India. What are the side effects of this

medicine? Is there any long, bad effect of this

medicine? Please guide us.”

� “My son’s health form was returned by his college

because he had not taken the PPD (Mantoux) test

within the past 12 months. He was born in

Honduras and exposed to TB. His tine test in

1992, as in earlier parts of his life, was positive.

His doctor said he would always test positive since

he was exposed. His doctor also said he has a clean

x-ray and is in perfect health (he is 18 and has not

been in Honduras since we adopted him at age

four). Is it not true that once you test positive you

will always test positive, as the doctor said? What

do I tell the school? I cannot imagine that after all

the hard work to get into the college they would

keep him out for a positive test—especially if he

will always test positive. Please help me. Give me

something to tell the college health department.”

� “I tested positive on the TB skin test but I heard

about a new blood test for TB and want to get it

done. They don’t have it here where I live and my

doctor doesn’t know anything about it. Where can

I get the blood test? I am willing to fly to any state

where it is available.”

EVOLUTION OF TREATMENT GUIDELINES

In 1965, treatment for latent TB infection was first

recommended for those with previously untreated TB,

TST converters and all children under age three with a

positive tuberculin test. The recommendations were

broadened in 1967 to include all who were TST-positive

(>10 mm) and their close contacts.

Treatment guidelines were developed in 1974

regarding pre-treatment screening and monitoring to

minimize the risk for hepatitis and exclusion of low-risk

persons over age 35 as candidates for treatment. The

guidelines were further revised in 1983 to recommend

INTRODUCTION


routine clinical and laboratory monitoring for persons

older than 35 or with increased risk for hepatoxicity.

In the years since HIV/AIDS was first reported in

1981, TB has reemerged as a tremendous threat to those

with compromised immune systems. Many of those

infected with HIV are co-infected with TB which is the

largest killer of HIV-infected persons worldwide. For this

reason, two months of Rifampin (RIF) and Pyrazinamide

(PZA) were recommended in 1998 for co-infected HIV-

positive persons, and subsequently for those who were

not HIV-infected.

Treatment guidelines for latent TB—what used to be

called “chemoprophylaxis”—have continued to be

refined. In 2000, nine months of Isoniazid (INH) was

decreed to be more effective than six months, and

months of RIF and PZA (2RZ) were deemed equal to nine

months of Isoniazid. But in 2003, 2RZ was de-

emphasized, due to its liver toxicity, in favor of nine

months of INH. Finally, 4 months of RIF was suggested

as being equal to the 9 months of INH regimen in efficacy

as well as with less toxicity.

TARGETED TUBERCULIN TESTING AND

TREATMENT OF LATENT TB INFECTION

As the rate of active tuberculosis in the United States

has decreased, identification and treatment of persons

with latent infection who are at high risk for active TB

have become essential components of the nation’s TB

elimination strategy.

In its 2000 report Ending Neglect: The Elimination of

Tuberculosis in the United States, the Institute of Medicine

(IOM) recommended precisely what this symposium is

intended to address: an increased emphasis on targeted

TB testing and treatment of latent TB infection. Among

its recommendations the IOM said:

� There should be increased emphasis on the

use of targeted TB testing and treatment of

latent TB infection. The focus should be on

identified groups that have a high incidence of TB,

including persons exposed to infectious cases, HIV-

positive individuals, persons born in high-

incidence countries, prisoners and other groups at

particular risk.

“To begin advancing toward the elimination of

tuberculosis,” said the IOM report, “aggressive new

efforts must be implemented to identify those who are at

greatest risk of disease through targeted programs of

tuberculin skin testing coupled with treatment for latent

tuberculosis infection.” It continued, “The question now

confronting the United States is whether another cycle of

neglect will be allowed to begin or whether, instead,

decisive action will be taken (1).”

Finally, the elimination of TB requires advances in

new diagnostic tools. While the advent of

QuantiFERON®-TB Gold In-Tube meets this objective,

continued advances are needed.

This monograph represents a step toward taking the

decisive action this country will need to eliminate the

ancient scourge of tuberculosis from America in the

21st century.

References

(1) Lawrence Geiter, ed. Ending Neglect: The Elimination

of Tuberculosis in the United States. Washington, D.C.:

Institute of Medicine, 2000.


Kitty Lambregts, M.D.,

Ph.D, M.P.H.

Dr. Lambregts is Senior

Consultant, International Unit

for the KNCV Tuberculosis

Foundation in The Hague,

Netherlands

There has not been much improvement in

tuberculosis testing since Green in 1951 spoke of

tuberculin in these mocking terms:

“It would surely simplify life for manufacturers if

Old Tuberculin were plainly described as any

witches’ brew, derived from evaporation of any

unspecified fluid medium in which any

unspecified strain of mammalian M. tuberculosis

had been grown, provided its potency matched

that of other witches’ brews, kept in Copenhagen

and called international standard, or any allegedly

equivalent substandard thereof, when tested on an

unspecified number of guinea pigs without

worrying too much about statistical analysis of

results” (1).

But as the tuberculin skin test is still an important

method for identifying infection with M. tuberculosis

in persons who do not (yet) have active disease, it

remains an important control-tool in the elimination

phase of tuberculosis.

EPIDEMIOLOGICAL SITUATION

As in all industrialized high-income countries, the

case rate in the U.S. is going down. The regular decline

in tuberculosis cases that resumed in 1993 reached an all

time low of 4.6 cases per 100,000 people (13,767 cases)

in 2006 (2). The U.S. has reached the elimination phase

of the TB epidemic (Figure 1).

As expected, case rates are not evenly distributed

across geographical areas, reflecting variations in both

sociodemographic and TB control situations. About 75%

of the new cases occur in the 99 metropolitan areas that

account for 62% of the total U.S. population. Although

the number of cases decreased 49% among individuals

born in the United States between 1992 and 1999, it

increased 2% among foreign-born persons. Individuals

from Mexico, The Philippines and Vietnam accounted for

nearly half of the foreign-born individuals with

tuberculosis (3). As a result the proportion of foreign-

born TB cases among all cases has increased and is likely

to continue increasing as the pool of infected Americans

becomes continues to shrink while the pool of infected

foreign-born persons becomes larger.

Figure 1

In some states, and nationally, the proportion of the

foreign-born with TB is larger than the proportion of TB

cases among patients born in the U.S. (Figure 2). High

percentages (>50%) of TB among the foreign-born are

seen in states with very low case rates (<3.5/100,000)

and in states with rates above the national average (2)

(Figure 3).

As expected, a relatively high proportion of TB in the

foreign-born is diagnosed within the first years of

immigration—reflecting recent infections in their

countries of origin. It is well known that progression from

infection to active disease is most likely to occur during

the first years of infection. However, TB cases will also

continue to occur among the growing pools of latently

infected foreign-born individuals, both legal and illegal,

who have resided in the U.S. for a longer period of time.

28,000

26,000

24,000

22,000

20,000

18,000

16,000

14,000

12,000

1982 1985 1990 1995 2000 2005

No

of C

ases

Year

Reported TB CasesUnited States, 1982–2006

RELEVANCE OF THE TUBERCULIN TESTIN LOW-PREVALENCE COUNTRIES


Figure 2

Figure 3

In addition to the category “immigrants from high

prevalence TB countries,” other risk groups for TB have

been identified in the U.S. as well. Tuberculosis is

relatively common among homeless people and in

individuals who reside in congregate facilities and

correctional institutions. Substance abuse also is

common in individuals with tuberculosis (3).

As in the U.S., The Netherlands has entered the

elimination phase of the TB epidemic without the use of

the BCG vaccination. The prevalence of TB infection in

consecutive age-cohorts in The Netherlands shows that

within two decades the pool of infected Dutch (excluding

foreign-born) will nearly have been eliminated (Figure 4).

The uneven distribution of TB in the world may, however,

influence this favorable situation to some extent.

Although most research indicates that transmission of

imported TB to the native population is limited (4, 5).

The extent of transmission, and thus the effect on

infection prevalence in the indigenous population, is

unknown and will probably vary. The effect of

importation on the TB situation in a country will depend

on the quality of TB control in that country and on the

magnitude and quality—prevalence of TB and drug

resistance—of the migration flow.

PERCENTAGE OF TB CASESAMONG FOREIGN-BORN PERSONS,

UNITED STATES

Lake Ontario

Lake Erie

Lake Huron

Lake

Michigan

Lake

Superior

FL

NM

DEMD

TX

OK

KS

NE

SD

NDMT

WY

COUT

ID

AZ

NV

WA

CA

OR

KY

ME

NY

PA

MI

VT NH

MARICT

VA

WV

OH

IN

IL

NC

TN

SC

ALMS

AR

LA

MO

IA

MN

WI

NJ

GA

WashingtonDC

Alaska

1996 2006

≥50%25%–49%≤25%

DCDC

1994 1996 1998 2000 2002 2004 2006

20000

15000

10000

5000

0

U.S.-born Foreign-born

NUMBER OF TB CASES IN U.S.-BORN VS FOREIGN-BORN PERSONs,

UNITED STATES, 1994–2006

Figure 4

In The Netherlands, three-quarters of all drug-resistant

TB cases are among the foreign-born. In 1997 and 1998

all MDR TB cases were associated with recent

immigration. In a cohort of 7,738 patients the prevalence

of INH resistance in asylum-seekers, regular immigrants

and Dutch citizens were respectively 10.3%, 7% and

2.8% (Table 1). Clearly, asylum-seekers and refugees

coming from unstable countries and war situations are at

increased risk of drug-resistant TB compared with regular

immigrants. In the United States, drug-resistant TB also

is associated with the foreign-born. Studies show high

overall rates of INH resistance (12%) and even higher

rates in large immigrant groups such as the Vietnamese

and patients from The Philippines (6). This is especially

relevant, as these high rates will reduce the overall

effectiveness of latent tuberculosis infection (LTBI)

treatment with nine months of INH in those populations,

a possible justification for adopting 4 months of rifampin.

Table 1

TB CONTROL CHALLENGES AND

OPPORTUNITIES

The positive effect of adequate TB control—i.e.,

decreasing incidence—at the same time hampers TB

control in countries in the elimination phase because

Rates of drug-resistance to INH (H), streptomycin (S)Rifampicin (R) and to HR in a cohort of 7,738 patients

diagnosed with bacillary tuberculosisin The Netherlands, 1993–1999

Resistance to

HSR

HR

Asylum seekers(n=1488)

%

10.39.92.61.7

Immigrants(n=2962)

%

77

1.50.9

Dutch(n=3288)

%

2.83.10.50.3


professional expertise and experience fade away.

Professional failure results in diagnostic delay, improper

use of diagnostic tools, inadequate treatment, poor

infection control and inadequate guidance of patients

(Figure 6). An analysis of TB diagnosis in The Netherlands

shows that a pulmonologist diagnoses on average 1.7 TB

cases a year. Laboratories diagnose an average of 16 new

culture-positive cases a year (Figure 5). At the same time

lack of interest among the general population and

policymakers results in patient delays, non-compliance

with treatment and lack of funds for TB control.

Figure 5

Figure 6

The combination of a “TB virgin population” and

diagnostic delay/improper treatment results in micro-

epidemics and concentration of TB in risk groups. An

analysis of well documented micro-epidemics in The

Netherlands shows that 1) micro-epidemics are associated

with Dutch nationality; 2) involve younger age-groups; 3)

occur all over the country, especially in lowest incidence

rural areas; and 4) are associated with long (often

combined) patient and doctor delays (Figure 6).

In addition to the waning expertise on TB, the

importation of (drug-resistant) TB is probably the

greatest challenge for industrialized countries (Figure 7).

Legal and illegal immigrants and especially those seeking

PROBLEMS AND CHALLENGES (1)

• Doctor’s/patient’s delay• Inadequate treatment• Inadequate infection control

• Micro-epidemics• Risk groups• Nosocomial infections

• Professionals

• Population

• Politicians ($)

Expertise

Disease incidence

{

DIAGNOSIS OF 6,511 PULMONARY TB CASESIN THE NETHERLANDS, 1993–1998

TB officer 2,383 37 35 11.3 Pulmonologist 3,422 53 344 1.7ID specialist 379 6 40 1.6Other 229 3 — —Unknown 98 1 — —

Cases Professionals Cases/year/ diagnosed registered professional N % N N

46 laboratories involved in mycobacteriology diagnose onaverage 16 culture positive cases/year/laboratory

asylum cause an unpredictable fluctuation in the TB

caseload. Transfers of immigrants in the country and

country-specific TB cultural stigma cause huge logistical

and transcultural problems and may thus hamper case

management. Mismanagement of these cases may lead to

prolonged transmission and spreading TB and drug

resistance. Another problem is that most immigrants are

not familiar with the concept of treating a disease that

does not cause symptoms, such as the treatment of LTBI.

The overlap of the HIV epidemic and the TB epidemic,

especially among substance abusers, homeless persons

and prisoners form a third important challenge for TB

control.

Figure 7

But some of the challenges mentioned above also offer

opportunities. After all, concentration of TB in certain

subpopulations allows for targeted interventions; These

interventions include: 1) targeted screening; 2) target

group-specific health education; and 3) target group-

specific treatment programs.

CONTROL INTERVENTIONS

TB control aims to break the chain of transmission by

(1) reducing the number of infectious sources in the

society by rapid diagnosis and adequate treatment of TB

cases, and (2) preventing the progression from infection

to disease by diagnosis and treatment of LTBI (Figures 8

and 9). There is a hierarchy here in that the identification

and treatment of active disease should have absolute

priority. Spending a lot of resources on LTBI treatment is

a waste if active tuberculosis is not adequately cared for.

This means that favorable conditions must exist for

passive case finding, including reducing the barriers to

care for all those suspected of carrying TB infection,

including illegal and/or uninsured individuals. In

addition, active case finding must focus on identified risk

groups for tuberculosis, such as recent immigrants and

PROBLEMS AND CHALLENGES (2)

Mismanagement anddefaulting

• Drug resistance• Transmission to Dutch

• Fluctuation caseload• Cultural barriers• Transfers

Import of TB


prisoners. Cohort analysis of treatment outcome provides

an essential element of TB control program monitoring.

High default and failure rates should lead to problem

analysis and appropriate action. In countries where

active cases are properly taken care of, diagnosis and

treatment of LTBI must be introduced to “speed up” the

elimination of TB.

This brings us to the role of the tuberculin skin test,

which is only of limited importance in source reduction

but of utmost importance as a tool in the “prevention of

breakdown” process (Figure 8). Treatment of LTBI

reduces the pool of latently infected individuals and thus

the pool of future TB sources.

Figure 8

Figure 9

In addition to the control interventions mentioned

above, infection control measures are increasingly

important. The overlap of the HIV and the TB epidemics

in some groups and settings requires that sufficient

measures be taken to prevent transmission from

identified and unidentified TB sources.

MONITORING THE TB EPIDEMIC

prophylaxis

BCG

early diagnosis

adequatetreatment

infected contact

tuberculin survey

DNA–fingerprinting

TB source

tuberculin survey: transmissionmolecular epidemiology: transmission resulting in disease

CHAIN OF TRANSMISSION—INTERVENTION

Preventionof

breakdown (II)

Preventionof

infection (I)

Infectedcontact

Source ofinfection

‡ KNCV 2002

THE ROLE OF THE TUBERCULIN SKIN TEST

(TST)Targeted tuberculin testing for LTBI is a strategic

component of TB control in low-prevalence countries

that identifies persons at high risk for developing TB and

who would benefit by treatment of LTBI. Persons with

increased risk for developing TB include those who were

recently infected and those who have clinical conditions

that are associated with an increased risk for progression

of LTBI to active disease (7).

Sometimes, for example, during contact investigations

or in patients with symptoms—a positive skin test result

(combined with x-ray and sputum examinations)

identifies patients who have already developed active TB.

In such cases the tuberculin skin test supports either

active or passive case finding rather than LTBI diagnosis.

Knowledge of tuberculin test sensitivity and

specificity, as well as understanding of the predictive

value of the test in different populations, are required to

properly target and interpret skin tests. False positive

tests occur in persons who have been infected with non-

tuberculous mycobacteria and those who have received

BCG vaccination. For this reason targeted testing should

only be conducted among groups at high risk and

discouraged in those at low risk. Causes for false

negative and false positive reactions are listed in Table 2.

It is important to distinguish the seven different roles

of the tuberculin skin test before discussing its relevance

and limitations. It is used for:

(1) Contact tracing

(2) Screening of exposure groups

(3) Screening of persons with clinical conditions

associated with progression to active tuberculosis

(4) Screening of risk groups for tuberculosis

(5) Screening of travelers to high-prevalence

countries

(6) Additional diagnostic tool in symptomatic

patients

(7) Measuring annual risk of infection/program

impact

Exposure groups are defined as “individuals who, by the

nature of their (voluntary) work, run a considerable risk of

being exposed to (unscreened) TB sources.” Risk groups are

defined (in Europe) as well-described sub-populations with

a TB incidence or prevalence of >100/100,000.


Table 2

As far as TB control impact is concerned, contact

tracing and screening of exposure groups are the most

important TST-related TB control interventions.

Contact tracing

In contact tracing, TST and QuantiFERON®-TB Gold

In-Tube offer great opportunities for TB control. A

confirmed contact with an infectious source strongly

increases the predictive value (PVP) of the test. Therefore,

a positive skin test result, and, moreso, a positive

QuantiFERON®-TB Gold In-Tube result, in contacts is

very likely to represent recent infection, and thus selects the

individuals who will benefit from LTBI treatment.

Diagnostic delay in TB sources may result in satellite cases

even before contact tracing is organized. In those situations

the TST selects individuals for x-ray and, therefore,

contributes to active case finding—but essentially comes

too late. However, timely diagnosis of symptomatic TB

patients and rapid initiation of contact investigations

usually allows contacts to be diagnosed with LTBI before

breakdown to active disease.

Contacts who have been exposed for more than two

months should not wait another eight weeks to get a TST

but should have one immediately and, if negative, a

second one 8-10 weeks after the last contact. In contrast,

contacts who were first exposed shortly before

identification of the source can wait for a single TST two

months later. But it is preferable to have a baseline TST

result to be able to document skin test conversion. All

contacts identified with LTBI should be offered LTBI

treatment. But in some cases—such as a 70-year-old

contact with liver disease—careful clinical monitoring

may be preferred.

The complexity of an extensive contact investigation

with different types of contacts requires specific skills,

understanding of TB pathogenesis and epidemiology and

central coordination.

CAUSES OF FALSE NEGATIVE OR FALSEPOSITIVE SKIN TEST

False negative

– BCG– MOTT

False positive– tuberculin used– administration–reading

biological – infections (not only HIV) – age – recent live virus vaccine – malnutrition – disease affecting T-cell function – drugs

technical

Exposure groups

In exposure groups, such as health care workers, it is

in the interest of the individuals involved to get a baseline

TST before exposure, which will serve as a reference

when these individuals are screened or take part in a

contact investigation after documented contact.

Whether members of exposure groups are periodically

screened or tested only after documented contact

depends on the local situation and risk assessment.

Documented skin test conversions should lead to

treatment of LTBI with recommended regimens (7). It is

of utmost importance that both the baseline and the

periodic/later TSTs are properly recorded for further

individual use and, as important, surveillance purposes.

TST-yield surveillance allows policymakers to judge the

necessity and consequences of TST surveillance in

specific groups. High levels of transmission, for instance,

should lead to evaluation of infection control measures

and, if found inadequate, measures for improvement.

Screening of persons with clinical conditions

associated with progression to active disease

Clinical conditions such as HIV-infection,

underweight, silicosis, diabetes mellitus, chronic renal

failure, gastrectomy, jejunoileal bypass, solid organ

transplantation and malignancies, and use of certain

drugs such as corticosteroids and other

immunosuppressive agents, increase the risk of breaking

down from LTBI to active disease (7). It needs to be

stressed that most of these clinical conditions may cause

false negative TST results (Table 2). But if LTBI is

diagnosed in these individuals, LTBI treatments should

be strongly considered. Patients with fibrotic lung

lesions, suggestive of inactive TB, should undergo careful

clinical examination (including TST and medical history)

and be offered LTBI treatment if active disease is

excluded by sputum examinations.

Risk groups for tuberculosis

In risk groups, such as immigrants coming from high-

prevalence TB countries, measures should first of all

focus on identification of active tuberculosis. TST can

serve as a selection mechanism for x-ray but, depending

on the target group, there may be different operational

and technical barriers—such as patients having to come

in twice, false positive reactions, boosting and false

negative reactions due to HIV or other conditions. This is

why x-ray screening may be preferred in most countries.


Secondly, LTBI treatment of individuals with

radiographic evidence of inactive TB is an important and

probably cost-effective intervention in tuberculosis

control programs in low-prevalence countries (8-10).

Obviously it is of great importance that active

tuberculosis is adequately ruled out.

Thirdly, treatment of LTBI in immigrants without x-

ray abnormalities may be considered in the elimination

phase of the TB epidemic. Recent experiences with

targeted outreach programs, using outreach workers of

the same cultural background as the individuals

diagnosed with LTBI—such as a program operating in

Seattle—indicate that excellent results can be obtained

provided (labor-intensive) LTBI treatment is tailored to

the individual needs of the patients (11).

But the aggressive approach of mandatory TST

screening and mandatory LTBI treatment of legal

immigrants as recently advocated by the Institute of

Medicine committee is debatable (3, 12).

Screening travelers to high-prevalence countries

Depending on the country of destination, duration of

stay and exposure to the indigenous population in that

country, tuberculin testing of travelers should be

considered. Testing should take place before leaving and

eight weeks after returning. Proper instructions to

travelers are crucial as compliance with the second test is

usually poor. LTBI treatment should be offered to all

converters after excluding active disease.

Skin testing of symptomatic patients

The role of TST in symptomatic patients—such as

patients with lung infiltrates or pleurisy—is limited and

should be regarded as an additional diagnostic tool, which

cannot confirm nor exclude TB diagnosis. As

symptomatic patients are often cared for by general

practitioners in the private sector without sufficient

knowledge of TB epidemiology and the limitations of the

TST (e.g., false negative and positive results),

misinterpretation is common. On the other hand, an 18

mm skin test result in a 14-year-old American-born

patient with pleurisy or hilar lymphadenopathy can be

very helpful, especially as many of these cases are not

bacteriologically confirmed.

Measuring annual risk of infection and program

impact

Population-based surveys in both high- and low-

prevalence countries have proved to be very informative

in documenting the annual risk of infection and trend of

the epidemic, and thus (indirectly) program impact. In

the Netherlands for many decades all Dutch army

recruits were tested as well as certain school populations.

But these surveys were discontinued as the extremely low

prevalence among schoolchildren and the establishment

of a professional army no longer allow for reliable and

representative conclusions. In high-prevalence countries

such as Tanzania, Kenya and Vietnam, tuberculin surveys

are still used to measure the trend of the TB epidemic and

the influence of HIV (13, 14).

CONSEQUENCES AND LIMITATIONS OF THE

SKIN TEST (TST) Diagnostic procedures should have consequences;

otherwise one should not use them. When TST is

involved, most guidelines mention only LTBI as “the

consequence” of a positive test. And obviously LTBI

treatment is the most important consequence as it

prevents progression to active disease. Guidelines for

LTBI treatment can be found in the CDC publication

“Targeted Tuberculin Testing and Treatment of Latent

Tuberculosis Infection (7).”

There are other “positive” consequences that are

usually neglected. First, a TST test result—whether

negative or positive—may add valuable information in

symptomatic patients (passive case finding). Second, a

test result just below the cutoff point for that specific

target group should provide guidance to both the patient

and his or her general practitioner and other health care

professionals involved by raising the level of suspicion so

they will think of TB if there are symptoms. Third, in

some instances—such as with liver disease or among the

elderly who most likely have long been infected—the

physician may decide not to start LTBI treatment but

instead to offer close monitoring and proper instruction

to the patient.

But a TST can only be used in an effective and cost-

effective way if those who decide to use it realize its

limitations. First of all, there may be specificity problems

due to cross-reactions with mycobacteria other than

tuberculosis and BCG vaccination. Second, there may be

sensitivity problems in the elderly and newborns or due

to infections—especially, but not only, HIV—or serious

diseases (including TB) (Table 2). Third, in contrast to

what is usually said, the TST is not at all an easy test!

Lastly, the most important consequence of a positive

test—namely, LTBI treatment—is not effective in

individuals with drug resistance to the drugs used for


LTBI treatment. This is why if there is documented

contact the LTBI regimen should be based on the drug

resistance pattern of the index patient.

False positive and negative reactions are due to either

technical or biological causes or both (Table 2). The

technical problems are generally underestimated.

Storage of tuberculin, how long one

can use a “prepared syringe,” the intracutaneous

administration and the reading are all elements of the

same test which require both knowledge and experience.

Below are some examples:

A study done among health care workers (medical

doctors and nurses) in a large teaching hospital in The

Netherlands found insufficient knowledge of TST

techniques (Table 3):

� 93% of subjects felt able to correctly perform the

TST

� 40% knew how to inject tuberculin

� 45% knew how to read the results

� 27% had sufficient knowledge to execute the whole

test procedure

Table 3

Targeted testing (testing for a valid reason) should be

the responsibility of those health care workers who have

been adequately trained.

Using the Mantoux skin test in a large population of

Netherlands army recruits, a substantial number of

those testing positive actually were co-infected with a

non-tuberculosis mycobacteria, M. scrofulaceum (15)

(Table 4). Also among a total of 237,692 Netherlands

army recruits between 1986 and 1993, 172 (48%) of the

355 individuals with induration diameter >10 mm

and <15 mm had presumed false-positive tests (15)

(Table 5). This is a typical example of non-targeted TST

Insufficient knowledge of TST-techniquesamong health care workers in a largeteaching hospital in The Netherlands

93% of subjects felt able to perform TST correctly

40% knew how to inject tuberculin

45% knew how to read the results

27% had sufficient knowledge to execute the whole test procedure

G.H. Poortman et. al. Parate Kennis over de uitvoering van de Mantoux-test onvoldoende NedTijdschr Geneeskd 1999 17 April:143(16)

(for surveillance purposes only) and illustrates why TST

should be targeted and why cutoff points in the U.S.

range from 5-15 mm depending on the predictive value

of the test in the target group (7).

Table 4

Table 5

Further complicating the use of TST is the problem of

boosting of tuberculin sensitivity. In one study of

Southeast Asian refugees, Veen found temporary anergy in

35% of 221 Vietnamese refugees. There was an association

with BCG history, though not with reactivity to sensitin or

M. scrofulaceum (16). Likewise, Cauthen et al. found that

30.9% of 2,469 initial non-reactors among 2,469 refugees

from South East Asia boosted on a subsequent TST.

Boosting was associated in this case with both reactivity to

MOTT sensitins and a BCG history (17).

Induration diameters to PPD RT 23between 10 mm–15 mm and percentage of

presumed false positives, 1986–1993:using the 3 mm criterion

1986 13,353 23 10 (43)

1987 12,380 21 8 (38)

1988 12,022 21 12 (57)1989 43,801 59 27 (46)1990 41,040 58 24 (41)1991 39,956 46 14 (30)1992 39,655 55 39 (71)1993 35,585 72 38 (53)Total 237,692 355 172 (48)

Year Persons No. with induration Presumed falsetested diameter 10 mm and 15 mm positive (%)

Source: Hans Bruins, PhD thesis: Mantoux skin testing andisoniazid prophylaxis in The Netherlands Army, September 1998

Reactivity to PPD-RT 23 and M. scrofulaceum sensitin in 37,755 army recruits without

previous BCG-vaccination 1986–1988,The Netherlands

1986 13,353 0.47 5.1

1987 12,380 0.41 7.9

1988 12,022 0.48 5.4

Year Persons PPD RT 23 M. scrofulaceum tested 10 mm (%) sensitin 10 mm (%)

Source: Hans Bruins, PhD thesis: Mantoux skin testing andisoniazid prophylaxis in The Netherlands Army, September 1998


CONCLUSIONS/SUMMARY ON THE ROLE OF

TST IN TB CONTROL

Elimination of tuberculosis is a public health

assignment, which requires a uniform coordinated

approach. Styblo proposed a definition of elimination as

a prevalence of infection in the general population less

than one percent and/or an incidence of smear-positive

tuberculosis of less than one per million (17).

The U.S. is ready for this challenge. Although case

rates are low, the U.S. is confronted with import of drug-

resistant tuberculosis and concentration of tuberculosis

in risk groups such as illegal immigrants, the homeless

and prisoners. At the same time these challenges offer

opportunities as they allow for a targeted approach.

TB control priority number one is the timely diagnosis

and adequate treatment of active TB cases. Priority

number two is the identification of persons who are

latently infected and run significant risk of progression to

active TB disease. High risk reactors are 1) contacts, 2)

skin-test converters and 3) persons with a clinical

condition associated with progression from LTBI to active

disease. Target groups for systematic TST are described

in the text. Tuberculin testing plays a limited role in TB

diagnosis in symptomatic patients as it does not allow for

exclusion or confirmation of TB diagnosis. Nevertheless,

in some clinical situations the skin test indeed adds

useful evidence.

Tuberculin testing should be conducted only among

groups at high risk for infection with M. tuberculosis and

discouraged in those at low risk. This targeted approach

relates to sensitivity and specificity problems of the TST

under low prevalence conditions. Knowledge of the

predictive value of the test in different groups is required

to properly use and interpret skin tests. False positive

tests occur in persons who have been infected with non-

tuberculous mycobacteria and those who have received

BCG vaccination. False negative results are associated

with both technical and biological conditions (infections,

underweight, diseases). The TST is not an easy test, so its

use should be limited to health care workers who are

properly trained.

But we need to stress that, especially with limited

public health resources available, targeted use of TST

should be introduced only after: (1) priority one (basic

TB control) interventions are put in place; (2) patients

diagnosed through passive case finding are adequately

treated; (3) an LTBI treatment program tailored to the

needs of the target groups has been designed; and (4)

cohort analysis of patients put on LTBI treatment is

introduced (treatment completion rate, default rate, etc.).

In addition to identifying LTBI, TST surveillance

serves to monitor transmission in exposure groups such

as health care workers.

In conclusion, despite all the limitations of the

tuberculin skin test it is an important tool in the

elimination phase of a TB control program provided: 1)

its use is targeted; 2) it is properly administered and

read; 3) interpretation is linked to “why” and “in whom”;

and 4) consequences reflect both individual and public

health interest.

Gustav Fischer offered a wise message in an 1891

booklet on the tuberculin test:

“Es können mithin in der Hand des erfahrenen

Arztes, welcher sich der Bedingungen der

Anwendungs—und Einwirkungsweise im

Einzelfall bewusst ist, die bacillären

Stoffwechselproducte zu Heilmitteln werden.”

(“In the hand of an experienced physician, who

realizes the conditions under which the test

should be administered and works, can this

bacillary product become a tool for cure.”) (18)

Although Fischer may have expected too much, he

was entirely right about the fact that it is crucial for

health care professionals involved to understand the

limitations and opportunities of TST.

References

(1) Green HH. Discussion on tuberculins in human

and veterinary medicine. Proc Roy Soc Med 1951;

44:1045.

(2) U.S. Centers for Disease Control and Prevention,

Atlanta, GA. <www.cdc.gov>.

(3) Lawrence Geiter, ed. Ending Neglect: The Elimination

of Tuberculosis in the United States. Washington,

D.C.: Institute of Medicine, 2000.

(4) McKenna MT, McGray E, Onorato I. The

epidemiology of tuberculosis among foreign-born

persons in the United States, 1986 to 1993. N Engl

J Med 1995; 332: 1071-76.

(5) Lillebaek T, Andersen AB, Bauer J, et al. Risk of

Mycobacterium tuberculosis transmission in a low-

incidence country due to immigration from high-

incidence areas. J Clin Microbiol 2001; 39: 855-61.


(6) Talbot EA, Moore, McCray E, Binkin NJ. Tuberculosis

among foreign-born persons in the United States,

1993-1998. JAMA 2000; 284:2894-900.

(7) Centers for Disease Control and Prevention.

Targeted tuberculin testing and treatment of latent

tuberculosis infection. MMWR June 9, 2000 (49)

(RR-06):1-54.

(8) Nolan CM, Elarth AM. Tuberculosis in a cohort of

Southeast Asian refugees. Am Rev Respir Dis

1988;137:805-09.

(9) Dasgupta K, Schwartzman K, Marchand R,

Tannenbaum T, Brassard P, Menzies D. Comparison

of cost effectiveness of tuberculosis screening of

close contacts and foreign-born populations. Am J

Respir Rev Crit Care Med 2000; 162: 2079-86.

(10) Menzies D. Tuberculosis crosses borders. Int J Tuberc

Lung Dis 2000; 4: 153-59.

(11) Personal communication with Dr. Charles Nolan

during North American Region 7th Annual

Conference, Vancouver, March 2002.

(12) Coker R, Lambregts-van Weezenbeek K. Mandatory

screening and treatment of immigrants for latent

tuberculosis in the USA: just restraint. Lancet 2001;

1:270-276.

(13) Menzies D. Tuberculin surveys—why? Int J Tuberc

Lung Dis 1998; 2(4):263-264.

(14) Bosman MCJ, Swai OB, Kwamanga DO, et al.

National tuberculin survey of Kenya, 1986-1990.

Int J Tuberc Lung Dis 1998; 2 272-280.

(15) Hans Bruins, Ph.D. thesis. Mantoux skin testing

and isoniazid prophylaxis in The Netherlands Army,

September 1998.

(16) Veen J. Microepidemics of tuberculosis: the stone-

in-the-pond principle. Tuberc Lung Dis 1992;

73(2):73-6.

(17) Cauthen GM, Snider DE Jr., Onorato IM. Boosting

of tuberculin sensitivity among Southeast Asian

refugees. Am J Respir Crit Care Med 1994;

149(6):1597-600.

(18) Fischer G. Ueber die physiologische Grundlage der

Tuberculinwirkung. Jena 1891.



Dr. Morris is Chief of the

Laboratory of Mycobacterial

Diseases and Cellular

Immunology in the Center for

Biologics Evaluation and

Research of the U.S. Food and

Drug Administration, Bethesda,

Maryland

While Robert Koch’s original tuberculins did not

cure consumption (as he had hoped), skin tests

using tuberculin preparations have become one of the

most widely used diagnostic tests for detecting infection

with Mycobacterium tuberculosis. More than a decade after

Koch’s initial studies, tuberculin skin testing remains an

important diagnostic aide for clinicians and a valuable

epidemiologic tool for public health officials.

WHAT IS TUBERCULIN?Tuberculin is a complex mixture of culture filtrate

components obtained from sterilized cultures of tubercle

bacilli. Two types of tuberculin are licensed for the

United States market: Old Tuberculin (OT) and Purified

Protein Derivative (PPD). OT Is licensed to be used in

multiple puncture devices while PPD is available for

intradermal injection (Mantoux test) as well as for

percutaneous injection using multiple puncture devices.

Old Tuberculin

OT is a crude culture filtrate concentrate prepared

from heat-sterilized M. tuberculosis broth cultures. Briefly,

OT is prepared by growing tubercle bacilli in synthetic

media for 6 to 8 weeks, heating at 100oC to kill the

organisms, evaporating the media to one-tenth of its

original volume, and then filtering to eliminate the

bacteria. Although this antigen preparation is commonly

used in veterinary medicine, the current use of OT as a

skin test reagent in humans in the U.S. is limited.

Tuberculin PPD

Purified Protein Derivative of tuberculin (PPD) is a

protein precipitate obtained from filtrates of sterilized

cultures of M. tuberculosis. The initial PPD preparations

were prepared by Florence Seibert in the 1930s by

precipitating M. tuberculosis culture filtrates with either

trichloroacetic acid (TCA) or ammonium sulfate. A

generalized schematic describing the manufacturing of a

tuberculin PPD Master Batch is shown in Figure 1.

Briefly, six- to eight-week-old M. tuberculosis cultures are

heated with steam for several hours, filtered and then

precipitated with ammonium sulfate or TCA. The

precipitation step decreases the amounts of nucleic acid

and carbohydrate in the preparation and increases the

relative protein content.

Figure 1 Preparation of a master batch oftuberculin PPD

This decreased carbohydrate content in PPD relative

to OT contributes to the reduced number of nonspecific

immediate reactions seen after PPD administration (1).

After extensive washing with buffer, the precipitate is

lyophilized and stored in the cold. These lyophilized

PPD preparations are extremely stable and can be stored

for decades without significant decreases in potency.

Most manufacturers formulate the final product by

� Culture M. tuberculosis for 6-8 weeks in

synthetic media

� Kill the organisms by heating with steam

� Filter to remove bacteria and large particles

� Precipitate with ammonium sulfate or

trichlororoacetic acid

� Recover the precipitate by centrifugation

� Dissolve in buffer containing stabilizer and

preservative

� Wash by ultrafiltration with buffer

� Lyophilize the precipitate

DEVELOPING A CLINICAL TUBERCULINTEST: TUBERCULIN CHARACTERISTICS,REACTIVITY AND POTENCY


diluting the lyophilized PPD in a PBS-buffer containing

Tween 80 and a preservative such as phenol. Tween 80

is added to reduce the adsorption of tuberculin to glass

and plastics and thus, to minimize any reduction in

potency of the final liquid product during storage.

Tuberculin PPD preparations contain a mixture of

components including proteins and small amounts of

nucleic acids and carbohydrates. Low to medium-sized

proteins are believed to be the most active antigenic

components of these preparations. Characterization of

the protein components of PPD has been difficult using

common protein fractionation procedures. Silver

staining of polyacrylamide gels has revealed that

numerous protein bands are present in PPD preparations

with the majority of the bands having molecular weights

in the range of 10,000 daltons (2).

United States Standard PPD preparation

The United States PPD Standard (PPD-S) was

prepared by Florence Seibert in 1941 and is presently

maintained by the FDA’s Center for Biologics Evaluation

and Research. Prior to the approval of any PPD lot for

clinical use, all new PPD preparations must be tested in

bioassays and shown to have potency equivalent to PPD-

S. In the United States, one Tuberculin Unit (TU) is

defined as 0.02 mg of PPD-S. The standard American 5

TU dose of commercial PPD preparations is that dose

which induces a reaction size equivalent to reactions

elicited by 0.1 mg of PPD-S.

THE IMMUNE MECHANISMS RESPONSIBLE

FOR SKIN REACTIONS TO TUBERCULIN

When individuals become infected with

mycobacteria, T-cells primarily in the regional lymph

nodes proliferate in response to the antigenic stimulus.

Within several weeks, these sensitized lymphocytes

circulate in the bloodstream, the injection of tuberculin

into the skin reticulates the sensitized lymphocytes,

which are subsequently responsible for the events

leading to the local reaction. The reaction is called a

delayed-type hypersensitivity (DTH) reaction because of

its delayed course; the reaction to tuberculin usually

begins at six to eight hours after administration of the

antigen, becomes maximal at 48-72 hours, and usually

wanes after several days.

The immune response to the tuberculin is initiated

when the sensitized lymphocytes release cytokines and

chemokines, which mediate the infiltration of other

immune cells into the site of antigen deposition. The

eventual dermal reactivity seen in positive responders

involves vasodilation, edema and the cellular infiltration

of lymphocytes, basophils, monocytes and neutrophils.

Only a small proportion of the cells at the site of dermal

reactivity are actually sensitized to mycobacterial

antigens; most of the cellular recruitment occurs in

response to the lymphokine release by the sensitized

T-cells. Sufficient T-lymphocyte sensitization to produce

positive DTH dermal reactions to tuberculin injections

usually occur two to ten weeks after infection with M.

tuberculosis. This dermal sensitivity often can be detected

several decades after the initial infection, although

frequently skin test reactivity wanes with advancing age.

CROSS-REACTIVE SKIN REACTIONS

Numerous studies in humans and animal models have

documented the cross-reactive nature of skin test

responses to tuberculins. For instance, some patients

with culture-confirmed M. avium disease react strongly to

tuberculin PPD while patients infected with

M. tuberculosis can elicit significant DTH responses to

MAC sensitins. In addition, the reactivity induced by

BCG vaccination confounds interpretation of tuberculin

skin testing in individuals immunized with live

BCG. It has been recognized for decades using

immunoelectrophoretic techniques that the cross-

reactivity results because of the antigenic similarities

among proteins produced by the different species of

mycobacteria. In recent years, modern genomic and

proteomic analyses have confirmed the presence of

antigenic homologs in various species of mycobacteria.

Jungblut et al. has shown using comparative proteome

analysis of M. tuberculosis and M. bovis BCG that at least

90% of the proteins in these strains are identical (3).

Comparative genomic sequencing analysis and DNA

microarray assessments have also demonstrated a high

degree of similarity among the genomes of the slow-

growing mycobacteria (4-5).

In addition to defining regions of identity in

mycobacterial genomes, the comparative molecular

analyses have elucidated genomic differences between

strains. For example, 16 genomic regions which are

present in M. tuberculosis are deleted in M. bovis BCG (4).

Based on these findings, the search for monospecific

diagnostic skin test antigens has intensified. Antigenic

proteins found only in M. tuberculosis such as Esat-6 and

CFP-10 are currently being evaluated as specific probes for

M. tuberculosis infections (6). Overall, these comparative

genomic and proteomic analyses have provided a novel


avenue for the rational design of specific mycobacterial

diagnostic reagents. It is likely the specific antigens or

combinations of specific antigens will be tested in the

clinic as monospecific skin test reagents in the near future.

THE STAGES OF CLINICAL DEVELOPMENT OF

SKIN TEST REAGENTS

The clinical development of most biological products

is initiated via pre-clinical studies in animal models and

is followed by clinical evaluations under the

Investigational New Drug (IND) system. If these studies

demonstrate that the products are safe and effective and

can be manufactured consistently, then a Biologics

License Application (BLA) can be submitted to the FDA

for review and possible approval. The four stages in the

clinical development of new skin test reagents are listed

in Table 1.

Table 1 The development of a skin testpreparation

I. Pre-clinical product development and

testing

• Develop manufacturing and testing

procedures

• Determine product safety in animal

models

• Assess product effectiveness in sensitized

animals

II. Investigational new drug stage

• IND submission and review of

manufacturing

• Phase I, II and III clinical trials

III.Biologics license application stage

• BLA submission and review

• Advisory panel recommendations

• Production facility inspection

• Bioresearch monitoring

• Product license approval

IV. Post-licensure stage

• Phase IV post-marketing clinical studies

• Adverse events reporting

• Lot release

I. PRE-CLINICAL PRODUCT DEVELOPMENT

AND TESTING

Pre-clinical testing in animal models should provide

insights into the product’s biological activity and safety as

well as help researchers to determine an optimal initial

formulation and select an appropriate starting dose for

clinical trials. Demonstration of safety is especially

crucial for any biological product being tested in the pre-

clinical phase. For tuberculin skin test reagents, safety

tests including general safety, sterility and freedom from

virulent mycobacteria assays should be completed prior

to clinical evaluation. The effectiveness and potency of

new skin test products should be estimated pre-clinically

using appropriate animal models.

II. INVESTIGATIONAL NEW DRUG STAGE

There are three distinct IND phases in the clinical

development of a new product (Table 2). Phase I

involves a safety study in a small number of people. For

a new tuberculin, a phase I study would test the safety of

the new product in 10-20 individuals. A phase II IND

study usually evaluates the safety and immunogenicity of

the product in a moderate number of subjects

(hundreds). Frequently, a phase II skin test study would

involve a dose-response assessment in which the dose of

the new skin test product that is bioequivalent to 5 TU of

the U.S. Standard is determined. At this stage, it is also

important to evaluate whether the skin test diluent

induces positive responses when administered alone.

The third IND phase, the basis for licensure, includes

pivotal trials to evaluate the safety and effectiveness in

larger number of individuals. A phase III trial for a new

tuberculin may compare the distribution of reaction sizes

to 5 TU of the U.S. Standard PPD with the bioequivalent

dose of the new product (as determined in phase II) in at

least three populations:

1. Persons known to be infected with M. tuberculosis

(sensitivity)

2. Persons living in areas with low mycobacterial

infection rates (specificity)

3. Persons living in areas with high rates of

nontuberculous mycobacterial infection

(specificity)


Table 2 Investigational new drug stage

III. BIOLOGICS LICENSE APPLICATION

STAGE

After successful completion of a phase II clinical

trial(s), the sponsor can submit a Biologics License

Application to the FDA for review. During this stage, the

clinical data is also submitted to an FDA advisory

committee for a recommendation on the safety and

effectiveness of the new product. In addition, the

manufacturing facility is inspected and bioresearch

monitoring of the clinical study sites is completed. If no

serious flaws in the product or the clinical data are

detected, then the BLA can be approved.

IV. POST-LICENSURE STAGE

After licensure, sponsors will frequently commit or be

encouraged to do additional studies to extend the

database for their product. For instance, the effectiveness

in a different patient population may be evaluated in

phase IV studies to extend the product labeling.

Extended safety monitoring of participants in phase III

clinical trials may be another post-licensure commitment

made by the manufacturer.

A critical component of the post-licensure stage is the

assessment of adverse events. After licensure, products are

administered to a larger population than has been tested

under IND and thus the detection of rare adverse events is

more likely. False-negative and false-positive reactions are

the predominant adverse events associated with tuberculin

testing. As shown in Table 3 (adapted from reference 1),

the factors that may cause false-negative responses include

co-existing infection or disease, improper skin testing

procedures, or underpotent product. False-positive

reactions have usually been associated with antigenic

cross-reactivity due to nontuberculous mycobacterial

infection or BCG vaccination. It is important for clinical

practitioners to report suspected false-positive or false-

negative tuberculin reactions to the manufacturer and to

the FDA. Reports can be submitted online to the FDA

through the MedWatch adverse event reporting program at

www.fda.gov/medwatch.

Following licensure, the FDA prior to their

distribution must approve the release of new commercial

lots for most biologic products. This lot release process,

which involves the review of testing results for the new

lot, helps to ensure that the product manufacture will be

consistent. For tuberculin, the manufacturer minimizes

lot-to-lot variation through the preparation of a large

Master Batch of lyophilized PPD. Since freeze-dried

tuberculin is extremely stable when stored appropriately,

clinical lots can be prepared (by dilution) from the same

Master Batch for several years. Guidelines for qualifying

a new Master Batch of PPD are provided in Table 4.

IND Phase I

� Safety testing in a limited number of individuals

IND Phase II

� Safety and immunogenicity

� Determine dose that is bioequivalent to 5 TU of the PPD-S in persons known to be infected

with M. tuberculosis

� Evaluate the reactivity of the diluent

IND Phase III

� Pivotal trials to evaluate safety and effectiveness

� Compare the distribution of reaction sizes to 5 TU of PPD-S with the bioequivalent dose of the new

skin test product in at least three populations:

– Persons known to be infected with M. tuberculosis

– Persons living in areas with low exposure to mycobacterial infection and presumed to be not infected with

M. tuberculosis or other mycobacteria

– Persons living in areas with high nontuberculous mycobacterial infection rates


Table 3 Potential causes of false-negativetuberculin reactions

Host factors

� Pre-existing infection

� Co-existing disease

� Nutritional deprivation

� Age of the person tested

� Stress

Product factors

� Improper storage

� Denaturation of the product

� Contamination of the product

� Use of expired product

Improper technique

� Inadequate tuberculin dose

� Injection too deep

� Errors in reading the skin test

� Reader bias

Similar to the IND process, the bioequivalence of lots

prepared from each Master Batch can be determined in

dose response studies using 5 TU of the U.S. Standard PPD

as a reference. The reactivity in the different target

populations listed in Table 4 should also be determined.

The bioequivalence of the human dose relative to the U.S.

Standard is then confirmed in guinea pigs. The formulation

of each commercial lot is based on the bioequivalency

determinations for the Master Batch. To ensure that each

commercial lot is correctly prepared, the potency of these

lots (relative to the U.S. Standard is determined in guinea

pig assays. Importantly, extensive safety testing is done on

each lot with general safety, sterility and freedom from

virulent mycobacteria assays being fundamental safety tests

for commercial tuberculin preparations.

Further information about this process can be obtained

from several sources. The regulations concerning current

good manufacturing practices and the clinical evaluation

of new products are listed in Title 21 of the Code of

Federal Regulations (sections 210, 312-314 and 600-610

are particularly relevant to biologic products). Information

about FDA guidelines, FDA points to consider documents,

and specific FDA standard operating procedures can be

obtained by visiting the FDA/CBER web site at

www.fda.gov/cber. Inquiries about the process can also be

directed to CBER’s Manufacturing Assistance and

Technical Training Branch at 800-835-4709.

Manufacturing

� A large Master Batch of tuberculin is prepared

� Clinical lots are prepared from the same Master Batch

Testing of the PPD Master Batch

� Standardization of the tuberculin against 5 TU of PPD-S

� Determination of the dose that is bioequivalent to 5 TU of the US Standard

� Comparison with PPD-S of the reactivity in tuberculous patients

� Comparison with PPD-S in areas with low rates of mycobacterial infection and in areas with high

rates of atypical mycobacterial infection

� Test the dose bioequivalent to 5 TU of PPD-S in guinea pigs. Compare laboratory potency results

to the results in humans

� Monitor stability

Lot release testing

� Assess the potency of new clinical lots in guinea pig assays

� Complete general safety, sterility and freedom from virulent mycobacteria assays

� Perform identity and purity tests

Table 4 Standardization of tuberculin PPD clinical lots


References

(1) Huebner RF, Schein MF, Bass JB. The tuberculin

skin test. Clin Infect Dis 1993; 17:968-975.

(2) Lachman PJ. Purified Protein Derivative. Springer

Seminars in Immunopathology 1988; 10:301-304.

(3) Jungblut PR, Schaible UE, Mollenkopf HJ, Zimmy-

Arndt U, Raupach B, Mattow J, Halada P, Lamer S,

Hagens K, Kaufman SHE. Comparative proteome

analysis of Mycobacterium tuberculosis and

Mycobacterium bovis BCG strains: towards functional

genomics of microbial pathogens. Mol Microbiol

1999; 33:1103-1117.

(4) Brosch R, Gordon SV, Pym A, Eiglmeier K, Garnier T,

Cole ST. Comparative genomics of the mycobacteria.

Int J Med Microbiol 2000; 290:143-152.

(5) Behr MA, Wilson MA, Gill WP, Salamon H,

Schoolnik GK, Rane S, Small PM. Comparative

genomics of BCG vaccines by whole-genome DNA

microarray. Science 1999; 284:1520-1523.

(6) Munk E M, Arend SM, Brock I, Ottendorf THM,

Anderson P. Use of ESAT-6 and CFP-10 antigens for

diagnosis of extrapulmonary tuberculosis. J Infect

Dis 2001; 183:175-176.


Elsa Villarino, M.D.,

M.P.H.

Dr. Villarino is Team Leader of

the Tuberculosis Trials

Consortium in the Clinical and

Health Systems Research Branch

of the Division of Tuberculosis

Elimination in the U.S. Centers

for Disease Control and

Prevention, Atlanta, Georgia

Editor’s Note: This article is an extension, expansion

and update of an article originally published in 1999

(Villarino ME, Burman W, Wang YC, Lundergan L,

Catanzaro A, Bock N, Jones C, Nolan C: Comparable

specificity of 2 commercial tuberculin reagents in persons at

low risk for tuberculosis infection. JAMA 1999; 281:169-

171), which addressed the possibility that one or both of the

commercially available PPD tuberculins might have an

unacceptably high rate of false positive reactions.

Context The possibility that one or both of the

commercial purified protein derivative (PPD) TB skin

testing products (Aplisol® and Tubersol®) may have an

unacceptably high rate of false-positive reactions.

Objective To compare the specificity and the distribution

of reaction sizes of Aplisol® and Tubersol®.

Design Randomized, double-blinded trial.

Setting Health departments (HDs) in Denver, CO;

Marion Co., IN; and Seattle-King County, WA; and

universities in Atlanta, GA; Tucson, AZ; and San Diego,

CA. The Denver and Marion Co. sites recruited primarily

HD clients or employees; all other sites primarily

recruited university students or employees.

Participants A total of 1,596 volunteers who, because

of their histories, were at low risk for infection with

Mycobacterium tuberculosis. Of these, 41 were excluded

from analysis (26 lost to follow-up, 15 ineligible owing to

various reasons).

Intervention Subjects received simultaneous skin tests

with four antigens: the standard PPD (PPD-S1), one each

of the two commercial PPDs and either a second PPD-S1

(24%) or PPD-S2 (a proposed new standard).

Main Outcome Measure Tuberculin skin test (TST)

reactions measured by two trained observers.

Results Using a 10-mm cutoff, the specificities of the

tests were equally high: Aplisol®, 98.2%; Tubersol®,

99.2%; and PPD-S1, 98.9%. Adverse events were minor

and not correlated specifically with any reagent. Using a

probability value of 0.05, no significant differences

between the 3 skin test reagents were detected in any of

the possible sources of skin-test variability analyzed: (a)

interobserver, (b) host and (c) reagent (differences

between PPD products and between unique lots of the

same product).

Conclusions With all other factors being equal, and

with a cutoff of at least 10 mm, either commercial PPD

reagent may be used with confidence for tuberculin

testing; both will correctly classify the same number of

uninfected persons.

INTRODUCTION

The development of antituberculosis drugs

revolutionized the treatment of tuberculosis (TB)

infection and disease, making TB both preventable and

curable. Preventability, nevertheless, depends first and

foremost on prompt and accurate diagnosis (1). Most

persons who are infected with Mycobacterium tuberculosis

never develop clinical illness, remaining asymptomatic and

non-infectious. However, latent infection can persist for

years and may progress to active disease. In the United

States an estimated 10 to 15 million persons are believed

to have latent tuberculous infection (LTBI) (2). Because a

large proportion of new active TB cases in the country

originate from this large pool of infected persons (3) it is

critical to diagnose infection and when appropriate, treat

the latent infection.

RANDOMIZED CLINICAL TRIALS OFSPECIFICITIES OF COMMERCIALLYAVAILABLE TUBERCULINS


The tuberculin skin test is the standard method for

diagnosing infection with M. tuberculosis (4). The skin

test involves intracutaneous injection of 5 tuberculin

units (TU) of purified protein derivative (PPD) by the

Mantoux technique. Standard PPD (PPD-S1) tuberculin,

lot no. 49608A, was prepared in 1940 and adopted in

1951 as an international standard for tuberculin PPD by

the World Health Organization Expert Committee on

Biological Standardization (5). Today the remaining

PPD-S1 antigen is stored and released for use by the Food

and Drug Administration (FDA). Master batches of

commercial PPD are standardized against PPD-S1 by

comparative testing in human populations with and

without an identified risk for infection with M.

tuberculosis. Individual lots of commercial PPD are

subsequently standardized against PPD-S1 only through

guinea pig potency tests. The standard 5-TU dose of

commercial PPD used in the United States should

produce reactions equivalent to PPD-S1 + 20%.

Two companies manufacture PPD tuberculin in the

United States: JHP Pharmaceuticals (Aplisol®) and

Pasteur Mérieux Connaught (Tubersol®). Despite FDA

regulations for production and standardization of PPD

tuberculin, there have been concerns that these

commercial PPD products vary in performance. Clusters

of unexpected positive reactions or suspected false-

positive results involving both products have been

reported in the medical literature, to the FDA (3) and to

the Centers for Disease Control and Prevention (CDC)

(6-12). The accurate diagnosis of infection is important

to ensure that infected persons receive appropriate

evaluation and treatment and that uninfected persons are

not exposed to unnecessary evaluation and treatment.

The possibility that one or both of the commercial

PPD products may have an unacceptably high rate of

false-positive reactions prompted our investigation. To

evaluate rates of false-positive reactions, it was necessary

to compare these products in persons who are unlikely to

be infected. To compare the specificity (the percentage of

uninfected persons correctly categorized) and the

distribution of reaction sizes of the two commercial PPD

reagents—Aplisol® and Tubersol®—we studied a

population of subjects who because of their history were

at low risk for infection with M. tuberculosis. We also

assessed the infection status of subjects included in our

study by simultaneous testing them with the “gold

standard” (PPD-S1) test.

METHODS

Study population

Study participants were recruited by investigators

from six sites: the Denver Public Health Department

(Denver), CO; Emory University (Atlanta), GA; the

Marion County Health Department (Marion Co), IN; the

University of Arizona (Tucson), AZ; the University of

California (San Diego), CA; and the Seattle-King County

Health Department (Seattle), WA. The persons solicited

for participation in Denver and Marion Co. were

primarily health department clients or employees; all

other sites primarily included university students or

employees. The Human Subjects Review Committee of

each site and of the CDC approved the study protocol.

All participants gave signed informed consent.

Eligibility criteria were (1) no risk factors for TB

exposure or infection with M. tuberculosis, as ascertained

by an eligibility questionnaire (available on request), (2)

age >18 years and <50 years and (3) birth in the United

States or Canada. Exclusion criteria were (1) known HIV

infection or other immunocompromising condition, (2)

previous immunization with BCG vaccine and (3) if

female, self-reported pregnancy. To evaluate the

immunogenicity of the antigens used for this study, we

used a second study population of persons expected to

be tuberculin reactors. Site investigators who had access

to TB case registries (all except Tucson) each recruited 20

persons who meet the following eligibility criteria: (1) a

history of bacteriologically-confirmed TB disease (within

five years from our study) and (2) a favorable clinical

response to at least 2 months of antituberculosis therapy.

All subjects in either study population who returned for

reading of the skin tests were paid for their participation

in the study.

Study materials

The skin test reagents used for this study were (1)

Tubersol® (lot numbers 2443-11 and 2458-11), (2)

Aplisol® (lot numbers 01206p and 00417p), (3) PPD-S1

and (4) PPD-S2. Two lots of each of the two commercial

products were used to simulate their availability in the

field, where at any given time more than one lot is

expected to be in use. The respective manufacturers

donated the commercial reagents. The FDA provided

PPD-S1 and PPD-S2. PPD-S2 was manufactured under

contract with FDA as a product bioequivalent to PPD-S1,

and is expected to replace PPD-S1 in the future as the

international tuberculin standard. The results of testing

with PPD-S2 are not included in this report (See:


Villarino ME, Brennan MJ, Nolan CM, et al. Comparison

testing of current (PPD-S1) and proposed (PPD-S2)

reference tuberculin standards. Am J Respir Crit Care Med

2000;161(4 Pt 1):1167-71). All skin test reagents were

injected using disposable plastic syringes with 28-gauge

needles (1 cc insulin syringes, Beckton Dickinson,

Franklin Lakes, NJ).

Randomization

Randomization lists were prepared for each of the six

study sites, using randomized blocks of antigen

sequences for groups of either three, six or nine patients.

Sequences were randomized by antigen and injection

site. Blocks were fixed so that approximately three

fourths of study subjects received an antigen sequence

that included PPD-S1 and PPD-S2, plus one each of

Aplisol® and Tubersol® (either lot); and one fourth of

the study subjects received an antigen sequence

including two separate injections of PPD-S1, plus one

each of either lot of Aplisol® and Tubersol®.

Skin testing procedures

Persons identified as eligible for the study were

interviewed, skin tested and given an appointment to

return for reading of the skin test at 48 or 72 hours after

the test. Information obtained by interview included

demographic, employment, residence and tuberculin

skin testing history. The four injections were placed on

the flexor surface of the forearm: two injection sites were

about two inches, and two injection sites were about four

inches below the elbow. Experienced study staff

following a standard protocol conducted the skin testing

and the reading at each study site. Two different persons

who were blinded to the identity of the test reagent and

to the other person’s readings read the skin-test results.

The results were recorded as the size in millimeters of the

transverse diameter of induration. The size of erythema

and adverse events were also recorded.

Sample size and statistical methods

Estimating a false-positivity rate of 4%, to detect with

80% power and 95% certainty a 2% difference between

the rates of false-positivity of Tubersol® and Aplisol®,

we estimated that the sample size needed was 1,146 (13).

To allow for potential errors in the assessment of

eligibility and losses to follow-up, we chose to enroll a

minimum of 1,500 healthy volunteers with low risk of

infection with M. tuberculosis.

We analyzed three potential sources of skin-test

variability in the low-risk study group. First, we assessed

the interobserver variability by comparing the reaction

sizes recorded by the two persons reading the same PPD-

S1 skin test. Second, we assessed the host variability by

comparing the reaction sizes recorded for the double

PPD-S1 skin tests. Third, we assessed the variability

between different antigens and between different lots of

the same antigen, by comparing the reaction sizes

recorded for each of the different skin test products. To

compare reaction sizes, we used nonparametric analyses

of variances (ANOVA), 95% confidence limits of the

differences among the mean reaction sizes, pair wise

Wilcoxon signed-rank tests and 0.05 probability values

adjusted for multiple comparisons (14-16). We examined

the interobserver correlation with a second method—the

Kappa statistic x 100(%)—that adjusts for chance

agreement between results measured by two different

persons (17). We also used nonparametric ANOVA, pair

wise comparison tests and probability values adjusted for

multiple comparisons to detect differences among PPD

reactions by the age, sex, race, place of birth and study site

of the study subjects (16-18).

We calculated the specificity of the skin test antigens

by two different methods. The first assumes that all the

low-risk subjects were truly not infected with M.

tuberculosis at the time they were skin-tested for this

study, and therefore that specificity equals 1 minus the

rate of reactions measuring >10 mm or >15 mm (false-

positive [FP] reactions) produced by testing with

Tubersol®, Aplisol® and PPD-S1. The second method

assumes that study subjects that had a >10 mm reaction

to tests performed with PPD-S1 were truly infected with

M. tuberculosis, and that therefore specificity equals one

minus the rate of FP reactions produced by testing with

Tubersol® or Aplisol®, presenting in persons who had

<10 mm reaction to tests performed with PPD-S1. Our

choice of cutoffs is based on the recommended values for

determining PPD skin test positivity for populations who

do not have an identified high risk for TB (2) (e.g., a

value of 10 mm is used for most health care workers

without other nonoccupational risk factors; the

recommended cutoff for most other low-risk persons is

15 mm).

To assess the capability of our study antigens to

perform as expected we analyzed the results of skin tests

given to persons who had culture-confirmed TB. For this

analysis we compared the mean reaction sizes of the three

different antigens with each other, as well as with the skin


test reaction sizes that have been recently reported for a

comparable population (Aplisol® mean=15.6 mm;

Tubersol® mean=15.0 mm) (19). We also assessed the

immunogenicity of our study antigens by calculating the

rate of false-negative reactions (reactions of <10 mm)

observed after testing with Aplisol®, Tubersol® and

PPD-S1 and compared these rates with the previously

reported average rate (10%) of nonreactivity to PPD in

patients with active TB (20, 21).

RESULTS

Study population

Between May 14, 1997 and October 28, 1997, we

enrolled and skin tested 1,596 persons with a low risk for

tuberculous infection. The demographic characteristics

of those excluded (n=41) did not differ from those

included (n=1,555) in the low-risk study group (Table

1). The reasons for exclusion included: failure to return

for reading (26), potential occupational exposure to M.

tuberculosis (6), birth outside the U.S. or Canada (5), age

>50 (3) and previous PPD-positive results (1). Of the

eligible low-risk subjects, results were read at 48 hours

for 1,417 (91%) subjects and at 72 hours for 138 (9%).

Ninety-nine persons with culture-positive TB were also

enrolled and skin tested; in these subjects results were

read at 48 hours for 93 and at 72 hours for 6. No person

from either study group experienced clinically significant

adverse reactions to PPD skin testing.

Interobserver variability

Included in this analysis are the results recorded by

two different persons who read the same PPD-S1 test in

each of 1,555 low-risk subjects. In this group, the mean

difference in paired reaction sizes for the replicate

PPD-S1 tests was 0.01 mm (range -14 mm to 9 mm) and

was not statistically different from zero (p=0.37). Among

127 persons in this analysis group who presented with

PPD-S1 reactions greater than zero, the difference in

reaction sizes recorded by the two observers was <2 mm

in 68 (54%) and >5 mm in 18 (14%). Using the Kappa

statistic and a cutoff value of 5 mm, there was a 75%

probability (due to reasons other than chance) that both

observers agreed on the test results when reading the

same PPD-S1 test, and a 72% probability that both

observers agreed on the test results when reading either

the same Aplisol® or the same Tubersol® tests. Based on

these results, for the remainder of the analysis and for the

three reagents, we used the average of the reaction size

readings recorded by the two observers.

Host variability

The number of low-risk subjects who received two

PPD-S1 tests was 360 and the mean difference in paired

reaction sizes observed between these two tests was 0.01

mm (range–11 mm to 13 mm). This difference is not

statistically different from zero (p=0.83). The difference

in reaction sizes was <2 mm in 22 (61%) and >5 mm in

4 (11%) of the 36 persons with double PPD-S1 tests that

presented with at least one PPD-S1 reaction greater than

zero. However, there were two (0.6%) situations in which

for the same person one of the PPD-S1 tests was read as

>10 mm and the other PPD-S1 test were read as <9 mm.

The mean reaction size observed among those subjects

who received only one PPD-S1 test (n=1,189) was 0.46

mm and it was not significantly different (p=0.98) than

the mean reaction size (0.27 mm) observed in those

subjects who received two PPD-S1 tests. Based on these

results, we used for the remainder of the analyses the

average of the two PPD-S1 reaction readings recorded for

each subject with double PPD-S1 tests.

Variability among different antigens

We compared and found no significant difference

between the mean reaction sizes obtained using the two

different lots of the commercial PPDs. Of the 1,555

persons in the low-risk study group, 765 (49%) were

tested with Aplisol® 01206p and 790 (51%) were tested

with Aplisol® 00417p. The mean reaction sizes

observed for the two Aplisol® lots were 0.58 mm and

0.56 mm respectively (p=0.77). The number of persons

tested and the mean reaction size observed with

Tubersol® lot 2443-11 was 792 (51%) and 0.40 mm;

with Tubersol® lot 2448-11, 763 (49%) and 0.30 mm

(p=0.91). For all other analyses we used the results of

testing with either lot of Aplisol® and Tubersol®.

Most (97%) skin test reactions to PPD-S1, Aplisol® and

Tubersol® measured <5 mm. The distribution of reaction

sizes varied significantly by study site (Table 2). This

variation was due to the larger rates at the San Diego site

among all size categories of reactions not equal to zero;

however, most of these reactions were small in size (263 of

288 [91%] ranged from 1-9 mm). For all study sites, the

rates for reactions >10 mm ranged from 0% to 3% for all

three antigens and were not significantly different, with the

exception of the San Diego site, where 4.9% of reactions

were >10 mm when testing with Aplisol®.


Table 1 Demographic characteristics of subjectsincluded and excluded in the low-risk oftuberculous infection study group

The ANOVA model comparing the mean reaction

sizes for Aplisol® (0.55 mm), Tubersol® (0.33 mm) and

PPD-S1 (0.40 mm) detected a significant difference

between at least two of the means (p=0.03). When we

examined the mean difference in paired reaction sizes

only the comparison between Aplisol® and Tubersol®

was statistically different from zero (p=0.007), with

Aplisol® producing larger reactions than Tubersol®.

The comparisons between the commercial reagents and

PPD-S1 did not show a significant difference (Aplisol®

vs. PPD-S1: p=0.22; Tubersol® vs. PPD-S1: p=0.14). We

found the same statistical similarities and differences

when we conducted the ANOVA and pairwise

comparisons in a group (n= 259) comprised of only those

low-risk persons who had at least one greater than zero

reaction to any of the three skin-test reagents. In this

group the mean reaction sizes to Aplisol®, Tubersol®

and PPD-S1 were 3.4 mm (SD=4.2 mm), 2.1 mm

(SD=3.2 mm) and 2.5 mm (SD=3.6 mm) respectively.

There were no significant differences in mean skin test

reaction sizes by age, gender, or race. However, there

Characteristic Included Excludedn=1,555 n=41

Median Age (range) 26 (18-50) 26 (18-58)

Male Sex (%) 590 (38) 18 (44)

Race/ethnicity (%)

White 1069 (69) 31 (74)

Black 209 (13) 6 (16)

Hispanic (all races) 180 (12) 3 (7)

Asian/Pacific Islander 50 (3) 1 (2)

American Indian/ 19 (1) 0AK native

Unspecified 28 (2) 0

Place of Birth* (%)

Western States 717 (46) 18 (44)

Central States 540 (35) 12 (29)

Eastern States 298 (19) 6 (15)

Not US or Canada 0 5 (12)

Student (%) 760 (49) 19 (46)

* Places of birth were grouped into geographic areas asfollows: WEST = WA, OR, ID, MT, WY, CA, NV, UT, CO, AZ,NM, AK, HI and Vancouver; CENTRAL = ND, SD, MN, WI,MI, NB, KS, IO, MO, OK, AR, IL, IN, OH, TN, KY, MS, AL, TXand LA; EAST = NY, CT, RI, MA, NH, VT, ME, PA, NJ, DL, WV,VA, DC, MD, NC, FL, GA, SC, Toronto and Montreal.

was a difference (p=0.0001) by study site. Subjects at the

San Diego site were more likely to have significantly

larger reactions (mean=1.3 mm) than persons from all

other sites (range=0.14 mm - 0.42 mm). There was also

a difference (p=0.0006) by place of birth. Persons born

in Western states had significantly larger reactions

(mean=0.6 mm), than persons born in either Central or

Eastern states (mean=0.3 mm for both). However,

further analysis showed that birthplace and study site are

not independent risk factors. Thirty-two percent of the

subjects born in the Western states were enrolled at the

San Diego site, and this association is significant

(p=0.004) when compared to the 20% enrolled by the

other two sites located in the West, and to the less than

3% enrolled by sites in Atlanta and Marion Co.

Test specificity results

For the first calculation of skin test specificity, all

reactions measuring >10 mm were considered false

positive (FP) reactions (Table 3). The number of FP

reactions and the test specificities calculated from 1,555

low-risk subjects tested were as follows: Aplisol®, 28

(98.2%); Tubersol®, 13 (99.2%); and PPD-S1, 17 (98.9%)

at the 10-mm cutoff; and Aplisol®, 7 (99.6%); Tubersol®,

2 (99.9%); and PPD-S1, 4 (99.7%) at the 15-mm cutoff.

In the second calculation of test specificity we excluded 17

(1.1%) of 1,555 subjects who were potentially true

positives on the basis of presenting with reaction sizes

measuring >10 mm after testing with PPD-S1. Based on

the skin-test results of the remaining 1,538 subjects, the

specificity of Tubersol® was 99.7% at the 10-mm cutoff

and 100% at the 15-mm cutoff. The specificity of

Aplisol® was 99.2% at the 10-mm cutoff and 99.7% at the

15-mm cutoff. There were no significant differences

between the specificities calculated by either of the two

methods.

Immunogenicity of study antigens

The reaction sizes observed in the group of 99 persons

with culture-positive TB had a mean of 16.2 mm (SD=5.8

mm) for Aplisol® a mean of 14.7 mm (SD=6.6 mm) for

Tubersol® and a mean of 16.1 mm (SD=6.2 mm) for

PPD-S1. There was no difference among the reaction-

size means of the three reagents using the ANOVA test,

and all the mean reaction sizes were similar to historical

controls. Thirteen persons in this group had reactions

measuring <10 mm to either PPD-S1 or Tubersol®.

Eleven persons had reactions <10 mm to Aplisol®.

These numbers are consistent with the previously

reported rate of nonreactivity to PPD in patients with TB.


Table 2 Reaction sizes observed after testing 1,555 low-risk subjects with PPD-S1, Aplisol® and Tubersol®,by study site

Table 3 Specificity and discrepant interpretations for Aplisol® and Tubersol® using two cutoff definitions

Number positive Specificity (%) p-value (x) Number positive Specificity (%) p-value (x)

PPD-S 17 98.9 4 99.7

Aplisol® 28 98.2 7 99.6

Tubersol® 13 99.2 0.02* 2 99.6 0.37Discrepant at 10 mm Discrepant at 15 mm

Number p-value Number p-value

A(+), PPD-S(-) 13 0.8 4

T(+), PPD-S(-) 5 0.3 0.9 0 0.12

* Pair wise comparison: PPD-S vs. A, p = .09; PPD-S vs. T, p = .35; A vs. T, p = .01

Atlanta Marion Co. Seattle San Diego Denver Tucson

PPD-S1

0 mm 241 229 211 185 239 294

1-4 mm 4 18 18 80 4 3

5-9 mm 2 2 5 13 3 2

>10 mm 4 1 1 6 4 1

Aplisol®

0 mm 237 228 208 183 236 288

1-4 mm 2 18 4 74 5 4

5-9 mm 5 2 7 13 6 7

>10 mm 7 2 1 14 3 1

Tubersol®

0 mm 241 234 214 196 243 295

1-4 mm 1 14 2 66 2 3

5-9 mm 4 2 4 17 2 2

>10 mm 5 0 0 5 3 0

Comment

This study shows that the reaction-size distributions

for Aplisol® and Tubersol® do not differ from those of

the standard PPD-S1 when these tests are applied

simultaneously in a population with a very low

likelihood of tuberculous infection. In our study

population, the specificity of both commercial products

was >98% using a cutoff for positivity of 10 mm. Noted

adverse effects were minor and not correlated in

particular to any specific reagent. Tubersol® produced

slightly smaller reactions, and Aplisol® produced slightly

larger reactions, than did PPD-S1. The larger reaction

sizes with the Aplisol® test may indicate that this

product contains more immunogenic material than the

other two reagents (22); however, the magnitude of this

difference appears to be of no practical public health

significance, since our estimated specificities for both the

Aplisol® and Tubersol® tests were similarly high.

We explored several potential sources of variability

related to the tuberculin skin test. The interobserver

agreement that we observed was very good; there was at

least a 75% probability that both readers agreed on the


associated with tuberculin skin tests not performed under

the same conditions. Moreover, a study that did conduct

simultaneous testing in 20 persons in Uganda did not

document any discordant results between the commercial

tuberculin reagents (24). Our study was randomized,

double-blinded and used simultaneous testing in a large

sample of well-characterized subjects with a low prevalence

of PPD positivity. We also included the results of skin

testing with the tuberculin standard, an assessment of the

interobserver and the host variability, and a verification of

the immunogenicity of the antigens used for the study, in

order to present as factually as possible our findings related

to the product differences between the two commercial and

the standard tuberculin reagents.

Skin test variation related to human factors can be

controlled only to a finite degree. In clinical practice, these

factors cannot be eliminated completely and should always

be recognized as potential sources of false-positive

tuberculin skin tests. Notably, our study results

demonstrate that the choice of commercially available

products for performing tuberculin skin testing is not an

element in the test’s variation. With all other factors being

equal, the use of either Aplisol® or Tubersol® for

tuberculin skin testing will result in the same number of

persons being correctly or incorrectly identified as infected

with M. tuberculosis.

The correct diagnosis of infection is important to ensure

that persons at risk for TB be evaluated to exclude active

disease and offered treatment for latent TB infection if

indicated. Also of importance is the effect of prior

probability of infection in the predictive value positive of

tuberculin skin-testing: persons with no identified risk of

infection with M. tuberculosis should not be screened with

the tuberculin skin test because of the decreased likelihood

that a test can accurately predict the presence of infection

in these persons. The use of PPD skin testing under these

circumstances may result in several adverse clinical and

public health consequences, including (1) initiation of

unneeded therapy with its potential for medication-related

toxicities, (2) unnecessary use of health care resources in

uninfected persons and (3) unnecessary epidemiologic

investigations in situations where clusters of false-positive

results are found.

When there is an unavoidable need to conduct

tuberculin skin test screening in low-risk populations

(e.g., to fulfill employment or school registration

requirements), we recommend performing tuberculin

skin testing with either of the two commercially available

PPD antigens and using the recommended cutoff value to

interpretation of the test results by reasons other than

chance. We explored the host variability by examining the

results of two PPD-S1 tests done on one subject. We found

a discordance of >5 mm between the two PPD-S1 tests in

11% of 366 subjects, but only two (0.6%) situations where

there was disagreement in the interpretation of the skin

test as positive with a 10-mm cutoff. This amount of

disagreement has been reported as inherent to tuberculin

skin testing, and compares favorably with the results of a

previous study that applied duplicate skin tests of the same

lot number of PPD to a sample of more than 1,000

persons, and found discordance between the two tests in

13% of subjects (23).

We examined our study population and believe that it

is representative of U.S. population groups with low

likelihood of infection with M. tuberculosis; the overall

rate of positive (>10 mm) reactions was 1%. We did

detect what appeared to be an association between

having any measurable reaction and enrollment by the

San Diego site. We believe that the observed variations

by site do not represent true differences, but rather a

difference in skin-test reading experience at the different

study sites. The readers from all sites were experienced

in PPD-skin testing before the start of our study;

however, for most this expertise was acquired by working

in TB clinics. The two persons at the San Diego site have

been working together as a team in a formally constituted

skin-testing clinic for 11 years. These persons apply and

read >10,000 skin tests per year, including not only

tuberculin skin tests but other types of intradermal skin

tests (e.g., Candida, mumps, coccidioidin), that have

expected reaction sizes smaller than those of PPD. We

hypothesize that the readers from other sites might not

have as much familiarity with very small skin-test

reactions and might have read and recorded these small

reactions as zero induration. The inclusion or exclusion

of the results from the San Diego site does not affect our

results and conclusions in any significant way.

Clusters of unexpected positive reactions or suspected

false-positive tuberculin skin-test results have been

reported in the medical literature. These reports include

situations in which clusters of positive tuberculin reactions

have been noticed in groups of low-risk persons tested with

Aplisol® and which, on subsequent testing with

Tubersol®, were believed to be clusters of false-positive

reactions (6-12). None of these reports involved testing

with the two commercial products simultaneously and thus

cannot exclude the possibility of false-negative reactions

associated with Tubersol®, or another kind of error


determine test positivity (10 mm for health care workers

or others with potential occupational exposure to TB, 15

mm for all the others). Following these guidelines will

result in the minimum number of persons erroneously

classified as infected and potentially exposed to

unnecessary therapy.

References

(1) American Thoracic Society. Diagnostic standards

and classification of tuberculosis. Am Rev Respir Dis

1990; 142:725-35.

(2) Centers for Disease Control and Prevention.

Screening for tuberculosis and for tuberculosis

infection in high-risk populations: Recommendations

of the Advisory Council for the Elimination of

Tuberculosis. MMWR 1995; 44 (RR-11):18-34.

(3) Ferebee SH. An epidemiologic model of

tuberculosis in the United States. Bull Natl Tuberc

Assoc 1967; 53:4-7.

(4) Huebner RE, Schein MF, Bass JB Jr. The tuberculin

skin test. Clin Infect Dis 1993; 17(6):968-975.

(5) Landi S. Production and Standardization of

tuberculin. In: Kubica GP and Wayne LG, eds. The

Mycobacteria. New York: Marcel Dekker, Inc., 1984,

pp 505-535.

(6) Blackshear JB, Exequiel B, Gesink D, Davies SF, Iber

C, Johnson JR. False positive skin tests with Parke-

Davis Aplisol®. Am Rev Respir Dis 1983; 127:254.

(7) Shands JW, Boeff D, Fauerbach L, Gutekunst RR.

Tuberculin testing in a tertiary hospital: product

variability. Infect Control and Hosp Epidemiol 1994;

15:758-760.

(8) Rupp ME, Schultz AW Jr., Davis JC. Discordance

between tuberculin skin test results with two

commercial purified protein derivative preparations.

J Infect Dis 1994; 169:1174-1175.

(9) Lamphear BP, Linnemann CC, Cannon CG. A high

false-positive rate of tuberculosis associated with

Aplisol: An investigation among health care

workers. J Infect Dis 1994; 169:703-704.

(10) Lifson AR, Watters JK, Thompson S, Crane CM,

Wise F. Discrepancies in tuberculin skin testing

results with two commercial products in a

population of intravenous drug users. J Infect Dis

1993; 168:1048-1051.

(11) Grabau JC, DiFerdinando Jr. GT, Novick LF. False

positive tuberculin skin test results. Public Health

Rep 1995; 110:703-706.

(12) Wurtz R, Fernandez J, Jovanovic B. Real and

apparent tuberculin skin test conversions in a group

of medical students. Infect Control Hosp Epidemiol

1994; 15:516-519.

(13) Pocock SJ. Clinical Trials. New York: John Wiley &

Sons, 1983.

(14) Lawless JF. Statistical Models and Methods for Lifetime

Data. New York: John Wiley & Sons, 1982.

(15) Westfall PH, Young SS. Resampling-Based Multiple

Testing. New York: John Wiley & Sons, 1993.

(16) SAS/STAT User’s Guide. Volume 1&2. Version 6, 4th

Edition. Cary, NC: SAS Institute Inc.

(17) Dean AG, Dean JA, Coulombier D, et al. Epi Info

Version 6: A word processing, database and

statistics program for epidemiology on

microcomputers. Atlanta: Centers for Disease

Control and Prevention, 1994.

(18) Fleiss JL. The Design and Analysis of Clinical

Experiments. New York: John Wiley & Sons, 1986.

(19) Duchin JS, Jereb JA, Nolan CM, Smith P, Onorato

IM. Comparison of sensitivities to two

commercially available tuberculin skin test reagents

in persons with recent tuberculosis. Clin Infec Dis

1997; 25:661-3.

(20) Nash DR, Douglass JE. Anergy in active pulmonary

tuberculosis. Chest 1980; 77:32-7.

(21) Holden M, Dubin MR, Diamond PH. Frequency of

negative intermediate-strength tuberculin sensitivity

in patients with active tuberculosis. N Engl J Med

1971; 285:1506-9

(22) Sbarbaro, JA. Skin test antigens: An evaluation

whose time has come. Am Rev Respir Dis 1978;

118:1-5.

(23) Chaparas SD, Vandiviere H, Melvin J, et al.

Tuberculin test: variability with the Mantoux

procedure. Am Rev Respir Dis 1985; 132:175-177.

(24) Johnson JL, Nyole S, Shepardson L, Mugerwa R,

Ellner JJ. Simultaneous comparison of two

commercial tuberculin skin test reagents in an area

with high prevalence of tuberculosis. J Infect Dis

1995; 171:1066-7.


ADMINISTERING AND READINGTUBERCULIN SKIN TESTS


Dr. Menzies is Professor in the

Departments of Medicine,

Epidemiology and Biostatistics

of the Montreal Chest Institute

at McGill University, Montreal,

Quebec, Canada

INTRODUCTION

Of the tests presently used in clinical medicine, the

tuberculin skin test is one of the few that was first

introduced in the 19th century. Given such a long history

of use, it may seem surprising that aspects of this test

remain controversial. The first tuberculin skin testing

material was developed by Robert Koch, who prepared it

by filtering heat-sterilized cultures of Mycobacterium

tuberculosis and then evaporating the filtrate to 10% of the

original volume (1). This became known as old tuberculin

(OT). Koch tried unsuccessfully to use this as a therapeutic

agent, but a few years later Von Pirquet described use of

the same material for detection of persons infected with

tuberculosis (2). In 1907, Mantoux introduced the

intradermal technique, which still bears his name (3).

Injecting tuberculin material intradermally into a

person previously infected with M. tuberculosis will result

in infiltration of previously sensitized lymphocytes from

circulating peripheral blood. At the site of the injection,

CD4 and CD8 T-lymphocytes, monocytes and

macrophages will accumulate. These release inflammatory

mediators, which produce edema and erythema.

Although this results in increased blood flow, the locally

increased metabolic activity of these inflammatory cells

results in relative hypoxia and acidosis, which may be

severe enough to lead to ulceration and necrosis (4).

Indications for tuberculin testing

Detection of latent TB infection should be done for

persons with the following conditions who have a high

risk of disease:

� HIV-infection

� Other immune compromised condition or therapy

• Cancer therapy

• Prednisone—20 mg/day or more for greater

than four weeks

• Other immune suppressants—e.g., TNF-αinhibitors (Infliximab), methotrexate

� Chronic renal failure, particularly if requiring

dialysis

� Diabetes mellitus

� Malnutrition

� Silicosis

Detect new TB infection (carries increased risk of

disease for subsequent two years)

� Contacts of active contagious TB case

� High risk of exposure:

• Occupational risk—work in health care, prison,

homeless shelters

• Travel to TB endemic areas (particularly if also

occupational risk)

Epidemiologic surveys and research

� Estimates of annual risk of infection as well as

prevalence of latent TB infection in different

population groups

� Research into factors associated with prevalent or

incident TB infection

ADMINISTRATION OF THE TUBERCULIN TEST

Tuberculin materials

At the present time the only accepted material for use

in tuberculin testing is purified protein derivative (PPD).

The production and standardization of this material is

described on pages 18-23 in this monograph.

Tuberculin test materials are commercially available in

strengths ranging from one (1) to 250 tuberculin units

(TU) per test dose. Administration of 1-TU is not

recommended because this preparation has sensitivity of

only 50% in children with confirmed active tuberculosis

(5) and 80% in adults with disease (6). Use of the lower

dose has been advocated to reduce occurrence of adverse

events, but there is no evidence that this dose is safer (7).

Higher strength formulations such as 100-TU or 250-TU

are not recommended because the resultant tuberculin


reactions will be much less specific. This is because

subjects sensitized to non-tuberculous mycobacteria will

react (8), with the result that reactions to the higher dose

test will not correlate with likelihood of true tuberculous

infection—as shown in Figures 1 and 2 (9). Therefore,

5-TU of PPD (equivalent to 0.1 microgram of PPD-S) is

strongly recommended.

Figure 1 Reactions to 5 T.U. and Contact with Tuberculosis

Figure 2 Reactions to 250 T.U. and Contact with Tuberculosis

Technique of test administration

The Mantoux method of intradermal injection is

recommended for administration of the test. The subject to

be tested should be seated with their arm supported

comfortably and the person administering the test should

also be seated comfortably facing the subject. The injection

should be made on the volar or flexor aspect of the forearm.

The site to be injected should be inspected; if there is

significant scarring, or there is eczema, other rash, or

infection then another site should be selected. If the

tuberculin test is repeated a few months after the first test,

it is important to change the site of injection. This is

because repeated injection at exactly the same site can

result in false positive reactions. The skin should be cleaned

and it is important to allow the cleansing solution to dry

before injection is made. The tuberculin material (0.1ml

equivalent to 5-TU) should be drawn up into a 1cc plastic

syringe no more than a half hour before administration.

Injection should be made with a 1/2-3/4 inch 26 or 27-

gauge needle. Use of smaller needles will result in less pain

and bruising and bleeding but may result in errors of

administration and less reliable results (10). When the

tuberculin material is injected, a small (5 mm diameter)

wheal should be created, although the size of the wheal

produced following intradermal injections correlates poorly

with the amount injected as it is affected by age and gender

(11, 12). If the injection is made too deep into the sub-

cutaneous tissue, then no such wheal will be seen. The

resultant reaction may be diffuse and harder to measure

resulting in false negative or false positive reactions (13). If

the injection is too superficial, the tuberculin material will

leak out on the skin, reducing the accuracy of the test.

bandages or dressings are not needed and not

recommended following tuberculin testing.

Other methods of test administration include multi-

puncture techniques such as the Tine® test (the Heaf test

was commonly used in Britain but has now been

discontinued). The Tine® test is popular as a screening

tool in pediatric populations. However, as shown in Table

1, multipuncture techniques are not as reliable as the

Mantoux, with lower sensitivity and specificity and are

not recommended (14).

READING THE TUBERCULIN TEST

Timing of reading can strongly affect results. Reactions

occurring after only 6 hours in one study was associated

with active disease—72% of 109 patients with smear

positive active TB had positive reactions after 6 hours

compared to only 3.5% of 143 healthy volunteers (15).

However, in subsequent studies, these early reactions

were found to be non-specific and so early readings are

not recommended. Compared to readings at 48-72

hours, readings after 24 hours have sensitivity of only

71% and a false positive rate of 9% (16). Reading after

seven days will also have lower sensitivity (6) particularly

in the elderly (17). Therefore, reading should be always

at 48-72 hours, i.e., on the second or third day.

The transverse diameter of induration should be

measured (12). Tuberculin reactions may exhibit

induration and erythema and, in some cases, the

erythema may be larger than the induration. It is

important to measure ONLY the induration—as

erythema has no relationship with the likelihood of true

TB infection. Originally induration was defined by


Table 1 Comparison of Mantoux with other tuberculin skin testing techniques

Author (Ref) Year Population Mantoux Dose Comparison Technique Age No. Type Sensitivity Specificity

(Mean or Range)

Badger (43) 1962 All ages 1001 5-TU Tine-OT 3mm 96% 73%6mm 78% 84%

Furcolow (44) 1966 36 670 5-TU Tine-OT 98% 92%100 95% 83%

Fine (45) 1972 54 589 5-TU Tine-OT 98% 82%

Wijsmuller (46) 1975 Adults 915 5-TU Jet injector 73% —

Donaldson (47) 1976 15-69 135 5-TU Tine-OT 84% —Tine-PPD 90% —

Lunn (48) 1980 18-21 250 5-TU Imotest – PPD 67% 65%(Merieux)

Ackerman (49) 1981 14 6239 10-TU Tine-PPD 76% —2574 Imotest-PPD 75% —

Hansen (50) 1982 Adults 829 5-TU Tine-OT 69% 98%

Rudd (51) 1982 Adults 100 10-TU Imotest-PPD 72% 94%

Biggs (52) 1987 Adults 105 10-TU Imotest-PPD 4mm 33% 90%2mm 60% 85%

palpation. The ballpoint technique was introduced by

Sokal (18) who considered this a faster, more reliable

technique. However in several studies, readings using the

two techniques have correlated very highly (19-22).

Nevertheless, the ball point technique appears to be

slightly faster (20), more sensitive (20) and less variable

(21).

It is important that the readings are made and recorded

in millimeters. It is NOT acceptable to record reactions

using terms such as “negative,” “doubtful,” or “positive.”

Rounding error or terminal digit preference is a common

problem with inexperienced readers, who tend to round

up or down to multiples of 5 mm. To minimize this

problem, readers should use simple machinist or tailors

calipers, which prevent reading the size in millimeters at

the same time the diameter is defined.

It is important that a trained health professional read

the test. When compared with health professional

reading, patient self reading has an unacceptably high

false negative rate. As shown in

Table 2, patients will often underestimate their reaction

when it is considered positive by a trained observer (16).

ADVERSE REACTIONS

Immediate wheal and flare with a local rash was

reported in 2.3% of patients in one series (23). These

reactions were associated with atopic history and were

not associated with positive tuberculin reactions at 48-72

hours—i.e., they were completely independent events.

Lymphangitis has been reported following tuberculin

testing, and is usually associated with large tuberculin

reactions (24). Anaphylaxis following tuberculin testing

has been reported in a total of three occasions. Of these,

only one occurred in a patient who was tested with the

Mantoux technique. This patient had active tuberculosis

of the lymph nodes, and developed shock with renal and

hepatic dysfunction hours after receiving a 1-TU

tuberculin test. The other two cases were associated with

Tine® testing, one of which was fatal (25, 26). In the

non-fatal case serum IgE to the tuberculin material was

not detectable and the authors believed that the gum

used as an adherent was responsible (25).

Table 2 Who should read the tuberculin test?

In approximately 1-2% of patients with positive

tuberculin reactions, there may be severe blistering and

Patient self-reading

Positive Negative TotalsHealth Positive 79 133 212professionalreading Negative 5 520 525


even ulceration. Cold compresses and anti-

inflammatories will provide symptomatic relief. If

blistering is present, cover with a dry dressing to prevent

scratching, which will break the blister and can lead to

localized infection. Hydrocortisone cream is often given,

but was of no benefit in the only randomized controlled

trial to assess this therapy (27).

There is no evidence whatsoever that tuberculin

testing poses any risk in pregnancy (28), nor that

pregnancy affects results.

FALSE NEGATIVE TUBERCULIN TESTS

As summarized in Figure 3, false negative tuberculin

tests can occur because of technical factors, although use

of proper technique as suggested previously, should

eliminate these problems. Biological causes are also

common and are not as easily controlled. The occurrence

of temporary false negative tuberculin tests with live

virus vaccination, or viral infections including measles,

mumps and mono-nucleosis is well described. Since the

anergy usually lasts only one to two months, the

tuberculin test should be re-scheduled in patients with

this history.

Figure 3 False Negative Tuberculin Skin Tests inpatients with active TB

Use of the tuberculin test for diagnosis of active

disease is discouraged for three reasons. First, a positive

tuberculin test has a very low predictive value because

tuberculosis occurs in population groups with high

prevalence of latent TB infection (elderly, minorities,

foreign born). The tuberculin skin test does not

discriminate between latent infection and active disease.

Therefore, most patients with undiagnosed pulmonary

disease and a positive tuberculin test will actually not

have active TB. Secondly, false negative tests are common

in patients with active TB particularly with more

advanced disease, as shown in Figure 3. Finally, the test

is not without risk in these patients—the only reported

serious adverse event following Mantoux testing

occurred in a patient with active TB.

HIV infection is a major and important cause of false

negative tuberculin tests. The likelihood that a tuberculin

reaction will be false negative increases markedly as the

CD4 count declines, as shown in Figure 4 (29-31).

Interestingly, as shown in Figure 5, even though fewer

HIV infected patients demonstrate any reaction to

tuberculin, of those that do, the frequency distribution

and median size of tuberculin reactions is the same as in

HIV negative populations (31-34). It appears that

progressive HIV infection, rather than producing a

gradual waning with smaller and smaller reactions,

causes tuberculin reactions to suddenly “turn off,” as if a

threshold of immune dysfunction was reached.

Figure 4 Tuberculin Reactions in HIV InfectedPatients with Active Tuberculosis

Figure 5 Effect of HIV Infection on TST Reactions

Another very important cause of false negative tests is

older age. In North American populations, the proportion

of a positive tuberculin test increases up to the age of 65

then declines thereafter. Although the proportion

demonstrating any reaction to tuberculin diminished with

older age, the size of reactions did not change (35).

Longitudinal studies have demonstrated reversion of


positive tuberculin tests to negative in elderly nursing

home residents (17, 36, 37). As with HIV infected patients,

tuberculin reactions in the elderly do not seem to fade but

rather “turn off”—suggesting that a threshold of age related

immune dysfunction is reached in such patients.

ANERGY TESTING

Anergy testing has been suggested for the assessment

of individuals with negative tuberculin tests. Among HIV

infected patients with negative tuberculin tests, the

incidence of active tuberculosis was higher in those who

were anergic compared to those who were TST negative

but not anergic (38-40). However, the appropriate

antigens for anergy testing are unclear (41) and in

individual patients results of anergy testing can be very

misleading (42). In addition, recent trials have

demonstrated that therapy with Isoniazid for HIV-

infected patients who are anergic is of no benefit, in

contrast to HIV-infected tuberculin positive patients. For

these reasons, anergy testing is not recommended in

tuberculin negative HIV infected individuals.

References

(1) Koch R. An address on bacteriological research. Brit

Med J 1890; 2:380-383.

(2) von Pirquet C. Frequency of tuberculosis in

childhood. JAMA 1907; 52:675-678.

(3) Mantoux MC. La voie intradermique en

tubercalinothérapie. Presse Med 1912; 20:146-148.

(4) Swanson Beck J. Skin Changes in the tuberculin

test. Am Rev Resp Dis 1979; 120:59-65.

(5) Murtagh K. Unreliability of the Mantoux test using 1

TU PPD in excluding childhood tuberculosis in

Papua New Guinea. Arch Dis Child 1980; 55:795-799.

(6) Duboczy BO, Brown BT. Multiple readings and

determination of maximal intensity of tuberculin

reaction. Am Rev Resp Dis 1960; 82:60-67.

(7) Spiteri MA, Bowman A, Assefi AR, Clarke SW. Life

threatening reaction to tuberculin testing. Brit Med J

1986; 293:243-244.

(8) Gryzbowski Stefan, Brown MT, Stothard D.

Infections with Atypical Mycobacteria in British

Columbia. CMAJ 1969; 100:896-900.

(9) Palmer CE. Tuberculin sensitivity and contact with

tuberculosis. Am Rev Tuberc 1953; 68:678-694.

(10) Flynn P, Shenep J, Mao L, Crawford R, Williams B,

Williams BG. Influence of needle gauge in Mantoux

skin testing. Chest 1994; 106:1463-1465.

(11) Comstock GW. False tuberculin test results. Chest

1975; 68(3):465-469.

(12) World Health Organization. The WHO standard

tuberculin test. Geneva: World Health

Organization, 1963.

(13) Rhoades EV, Bryant RE. The influence of local

factors on the reaction to tuberculin. 1. The effect of

injection. Chest 1980; 77(2):190-193.

(14) Chaparas SD. Multiple Puncture Tuberculin Tests.

Pediatr Infect Dis J 1987; 6(5):496-497.

(15) Glenchur H, Fossieck BE, Silverman M. An

immdediate skin test for the diagnosis of active

pulmonary tuberculosis. Am Rev Resp Dis 1965;

92:741-748.

(16) Howard TP, Solomon DA. Reading the tuberculin

skin test: who, when and how? Arch Intern Med

1988; 148:2457-2459.

(17) Slutkin G, Perez-Stable EJ, Hopewell PC. Time

course and boosting of tuberculin reactions in

nursing home residents. Am Rev Resp Dis 1986;

134:1048-1051.

(18) Sokal JE. Measurement of delayed skin-test

responses. New Engl J Med 1975; 293:501-502.

(19) Bouros D, Maltezakis G, Tzanakis N, Tzortzaki E,

Siafakas N. The role of inexperience in measuring

tuberculin skin reaction (mantoux test) by the pen

or palpation technique. Respiratory Medicine 1992;

86:219-223.

(20) Jordan TJ, Sunderam G, Thomas L, Reichman LB.

Tuberculin reaction size measurement by the pen

method compared to traditional palpation. Chest

1987; 92:234-236.

(21) Longfield JN, Marsileth AM, Golden SM, Lazoritz S,

Bohan S, Cruess DF. Interobserver and method

variability in tuberculin skin testing. Pediatr Infect

Dis 1984; 3:323-326.

(22) Bouros D, Zeros G, Panaretos C, Vassilatos C,

Siafakas N. Palpation vs pen method for the

measurement of skin tuberculin reaction (Mantoux

test). Chest 1991; 99:416-419.

(23) Tarlo SM, Day JH, Mann P, Day MP. Immediate

hypersensitivity to tuberculin in Vivo and In Vitro

studies. Chest 1977; 71(1):33-37.

(24) Morrison JB. Lymphangitis after tuberculin tests.

Brit Med J 1984; 289:413.

(25) Wright DN, Ledford DK, Lockey RF. Systemic

and local allergic reactions to the tine test

Purified Protein Derivative. JAMA 1989; 262(21):

2999-3000.


(26) DiMaio VJM, Froede CRC. Allergic reactions to the

tine test. JAMA 1975; 233(7):769.

(27) Hanson ML, Comstock GW. Efficacy of

hydrocortisone ointment in the treatment of local

reactions to tuberculin skin tests. Am Rev Resp Dis

1968; 97:472-473.

(28) Snider DE. The tuberculin skin test. Am Rev Resp Dis

1982; 125:108-112.

(29) Graham NMH, Nelson KE, Solomon L, Bonds M,

Rizzo RT, Scavotto J et al. Prevalence of tuberculin

positivity and skin test anergy in HIV-1-seropositive

and -seronegative intravenous drug users. JAMA

1992; 267(3):369-373.

(30) Huebner RE, Schein MF, Hall CA, Barnes SA.

Delayed-type hypersensitivity anergy in human

immunodeficiency virus-infected persons screened

for infection with Mycobacterium tuberculosis.

Clinical Infectious Diseases 1994; 19:26-32.

(31) Markowitz N, Hansen NI, Wilcosky TC, Hopewell

PC, Glassroth J, Kvale PA et al. Tuberculin and

anergy testing in HIV-seropositive and HIV-

seronegative persons. Ann Intern Med 1993;

119:185-193.

(32) Gourevitch MN, Hartel D, Schoenbaum EE, Klein

RS. Lack of association of induration size with HIV

infection among drug users reacting to tuberculin.

Am J Respir Crit Care Med 1996; 154:1029-1033.

(33) Johnson MP, Coberly JS, Clermont HC, Chaisson RE,

Davis HL, Losikoff P et al. Tuberculin skin test

reactivity among adults infected with human

immunodifiency virus. J Infect Dis 1992; 166:194-198.

(34) Okwera A, Eriki PP, Guay LL, Ball P, Daniel TM.

Tuberculin reactions in apparently healthy HIV-

seropositive and HIV-seronegative women - Uganda.

MMWR 1990; 39:638-646.

(35) Battershill JH. Cutaneous testing in the elderly patient

with tuberculosis. Chest 1980; 77(2):188-189.

(36) Gordin FM, Perez-Stable EJ, Flaherty D, Reid ME,

Schecter G, Joe L et al. Evaluation of a third

sequential tuberculin skin test in a chronic care

population. Am Rev Respir Dis 1988; 137:153-157.

(37) Perez-Stable EJ, Flaherty D, Schecter G, Slutkin G,

Hopewell PC. Conversion and reversion of

tuberculin reactions in nursing home residents. Am

Rev Resp Dis 1988; 137:801-804.

(38) Guelar A, Gatell JM, Verdejo J, Podzamczer D,

Lozano L, Aznar E et al. A prospective study of the

risk of tuberculosis among HIV-infected patients.

AIDS 1993; 7:1345-1349.

(39) Moreno S, Baraia-Etxaburu J, Bouza E, Parras F,

Perez-Tascon M, Miralles P et al. Risk of developing

tuberculosis among anergic patients infected with

HIV. Ann Intern Med 1993; 119:194-198.

(40) Antonucci G, Girardi E, Raviglione MC, Ippolito G,

for the GISTA. Risk factors for tuberculosis in HIV-

infected persons. A prospective cohort study. JAMA

1995; 274(2):143-148.

(41) Pesanti EL. The negative tuberculin test. Tuberculin,

HIV and anergy panels. Am J Respir Crit Care Med

1994; 149:1699-1709.

(42) Chin DP, Osmond D, Page-Shafer K, Glassroth J,

Rosen MJ, Reichman LB et al. Reliability of anergy

skin testing in persons with HIV infection. Am J

Respir Crit Care Med 1996; 153:1982-1984.

(43) Badger TL, Breitwieser ER, Muench H. Tuberculin

tine test: Multiple-puncture intradermal technique

compared with PPD-S, intermediate strength. Am

Rev Resp Dis 1963; 87:338-351.

(44) Furcolow ML, Watson KA, Charron T, Lowe J. A

comparison of the tine and mono-vacc tests with the

intradermal tuberculin test. Am Rev Resp Dis 1967;

96:1009-1027.

(45) Fine MH, Furcolow ML, Chick EW, Bauman DS,

Arik M. Tuberculin skin test reactions. Am Rev Resp

Dis 1972; 106:752-758.

(46) Wijsmuller G, Snider DE. Skin testing: a

comparison of the jet injector with the Mantoux

method. Am Rev Respir Dis 1975; 112:789-798.

(47) Donaldson JC, Elliot RC. A Study of Co-positivity of

Three Multipuncture Techniques with Intradermal

PPD Tuberculin. Am Rev Resp Dis 1978; 118:843-846.

(48) Lunn JA, Johnson AJ, Fry JS. Comparison of

multiple puncture liquid tuberculin test with

Mantoux test. The Lancet 1981; (i):695-698.

(49) Ackerman-Liebrich U. Tuberculin Skin Testing.

Lancet 1982; ii(October 23):934.

(50) Hansen JP, Falconer JA, Gallis HA, Hamilton JD.

Inadequate sensitivity of tuberculin tine test for

screening employee populations. J Occup Med 1982;

24(8): 602-604.

(51) Rudd RB, Gellert AR, Venning M. Comparison of

Mantoux, Tine and ‘Imotest’ Tuberculin Tests. Lancet

1982; 515-518.

(52) Biggs B, Connor H, Dwyer BW, Speed BR.

Comparison of a multiple puncture tuberculin test,

“Imotest” and the Mantoux test in an Australian

population. Tuberc 1987; 68:285-290.



Dr. Menzies is Professor in the

Departments of Medicine,

Epidemiology and Biostatistics

of the Montreal Chest Institute

at McGill University, Montreal,

Quebec, Canada

INTRODUCTION

The use of repeated tuberculin tests to detect new TB

infections in high-risk populations has often

resulted in problems of interpretation. This is because

tuberculin reactions may change size because of random

variation of the test, or because of a real biologic increase—

and real increases may be due to boosting or conversion.

Random or chance variation

As shown in Table 1, the most important cause of

random variation in tuberculin test response is differences

in reading. Differences between readers result in standard

deviations of readings of 2.3 mm (1) or 2.5 mm (2).

When the same person re-reads the tuberculin reaction,

variability is even less and results in diagnostic test

misclassification of less than 2% (3). When multiple

tuberculin tests are administered and read, the resultant

test-to-test variation will include differences due to

administration and reading, as well as the inherent

biologic variability. The latter appears to be a small

effect—because the overall standard deviation is less than

3 mm—not much more than the differences in

administration and reading. Therefore, in 95% of

subjects, test-to-test variation due to biologic variability

plus differences in administration and reading should

result in differences of less than 6 mm (representing two

standard deviations). This is why 6 mm is generally used

as the minimum criteria in order to distinguish a true

increase in size from that due to chance variability alone.

Although rational, use of these criteria in day-to-day

management can sometimes present difficulties. For

example, if an individual has a first tuberculin reaction of

8mm and a second of 11 or 12 mm this could be due to

random variability. Yet in many populations, this

reaction would now be considered a positive test result.

The most pragmatic approach is to stop all further

tuberculin testing and ensure that the patient undergoes

radiographic and medical evaluation. Given the

likelihood that this now “positive” test may be only the

result of random test variability, the risk of tuberculosis is

likely to be low—if no other risk factors can be

identified. Therefore, the need for treatment of latent

infection should be correspondingly low.

The booster phenomenon

Boosting is defined as an increase in tuberculin

reactions of at least 6mm following repeat tuberculin

testing—that is, unrelated to new mycobacterial

infection. This phenomenon is believed to occur when

cell mediated response has waned resulting in an initially

negative tuberculin reaction, but the tuberculin test

stimulates anamnestic immune recall. When this

happens, a second tuberculin test, administered one

week to one year later evokes a much greater response.

The booster phenomenon was first described

following repeated tuberculin tests administered one year

apart. Subsequent studies demonstrated that the effect

was maximal if the two tests were separated by one to

four weeks, and were much less common if the interval

was less than 7 days.

As shown in Table 2, positive two-step second test

reactions are common in many populations and are

roughly correlated—although generally lower—with

prevalence of initial tuberculin reactions. For this reason,

the boosting phenomenon is common in the elderly

(4-8) and foreign born (9-13).

BCG vaccination has been consistently shown to have

an important effect on boosting (Table 3a), particularly

when the interval between vaccination and testing is

relatively short (14-16). BCG vaccination in infancy is

associated with less frequent boosting compared to those

vaccinated at an older age, also shown in Table 3b (17).

BCG can be associated with increased reactions of 25mm

or more as shown in Figure 1.

INTERPRETING REPEATEDTUBERCULIN SKIN TESTS


Another important cause of boosting is sensitivity to

non-tuberculous mycobacteria (NTM). As shown in

Table 4, in two studies individuals with sensitivity to NTM

antigens have a much higher occurrence of boosting with

tuberculin antigens even though the sensitivity to NTM

had minimal effect on the initial tuberculin reactions (17,

18). Given that the boosting phenomenon is associated

with remote TB infection, sensitivity to NTM antigens and

BCG vaccination, it appears to be a non-specific

manifestation of any prior mycobacterial infection.

The risk of later reactivation to active tuberculosis in

individuals with negative initial tuberculin tests but positive

second tests has been studied in only one prospective study

as shown in Figure 2. The risk of active TB in those with a

negative initial, but positive (10+ mm) second test was

approximately half the risk of subjects from the same

population with positive initial TST, consistent with the

finding that this phenomenon is non-specific (19).

Figure 1 Effect of BCG vaccination on changes inreaction size in Canadian-born workershaving two-step tests (PPD-T2–PPD-T1)

Table 1 Variability of tuberculin test results – (Mantoux test only)

Notes: *Patient self-reading compared to trained health professional

Variability of reactions: Two tests in the same subject

Author (Ref) Year Population Reading

No. Type Age Standard Deviation Misclassification

(Pos vs Neg)

Furcolow (3) 1966 212 Mental hospital 36 0-4 mm 92%

5-9 mm 7%

10+ mm 1%

Chaparas (28) 1985 1036 General Adults 4.6%

46 TB Patients Adults 0

Variability of readings – within readers

Reading

No. of Standard Misclassification

Readers Deviation (Pos vs Neg)

Bearman (29 1964 36 General 16-17 4 1.3-1.9 mm

Furcolow (3) 1966 670 Mental Hospital 11-90 2 1.2%

Patients

Variability of readings – between readers

Loudon (30) 1963 53 Workers 20-60 7 9%

Fine (31) 1972 189 General 54 4 12%

Erdtmann (2) 1974 121 General 18-25 4 2.5 mm

Perez-Stable (1) 1985 537 Nursing Home 770 6 2.3 mm 4.3%

Residents

Howard (32) 1988 806 General Adults 2* 11%

Pouchot (33) 1997 96 Health workers Adults 2 2.7-3.5 12-23%


Table 2 Prevalence of positive initial and second TST from two-step testing

Notes:† Initial test-% based on number undergoing T1, considered positive if ≥10 mm*Second test-% based on number undergoing T2, considered positive if: 0 = No definition given; 1 = T2 ≥ 10 mm, and T2-T1 ≥ 6 mm; 2 = T1<5, T2 ≤ 5 mm;3 = T1 < 10, T2 ≥ 10 mm ‡Number extrapolated from figures given in paper

Population Author (Ref) Setting No. subjects Percent with positive

undergoing T1

Initial test† Second test*

Health care workers Valenti(4) Rochester, NY 416 3.1% 01

Bass (34) Alabama N/A 8.2% 8.3%1

Menzies (17) Montreal 951 2.2% 2.5%1

Gross (35) Maryland 2558 3.8% 0.3%3

Richards (18) Chicago 267 2.6% 6.6%1

Nursing home residents Slutkin (36 San Francisco 411 35% 6%1

Gordin (5) San Francisco 1726 28% 14%1

Washington

W. Virginia

Alvarez (6) Tennessee 510 30% 19%0

Barry (8) Boston 323 26% 6%1

Van den Brande (7) Holland 223 29% 14%1

Hospital patients Burstin (37) Boston 162 12% 6%1

HIV-infected Webster (38) USA 709 N/A 2.7%2

Hecker (39) Uganda 345‡ 71% 29%2

IVDU Lifson (40) USA HIV- 900 13% 12%2

HIV+95 13% 8%3

Foreign-born Cauthen (41) USA 2469 36% 31%0

Menzies (11) Montreal 323 32% 16%1

Morse (12) N.Y.State 524 46% 43%1

Veen(13) Netherlands 221 39 18%3

The definition of boosting has been the subject of some

discussion; however, the only evidence regarding

prognosis of the booster reaction determined risk of TB in

those whose boosting reaction was defined as a second

test of only 10+ mm (19). Among Canadian health

workers, 5-9 mm two-step reactions were significantly

associated with the same clinical and demographic

characteristics as larger boosted reactions including older

age, foreign birth and BCG vaccination. This is also shown

in Figure 3. In populations such as health workers who

undergo two-step testing prior to periodic tuberculin

screening, it is suggested that individuals with small

reactions should also be considered to have the boosting

phenomenon, receive a medical and radiographic

evaluation and not tuberculin tested in the future. This is

because with subsequent testing, these individuals are

very likely to have further boosting, which would be

misdiagnosed as tuberculin conversion. It is unlikely that

the tuberculin test will be as useful or precise as is in other

individuals whose first and second tuberculin tests are less

than 5 mm.

Figure 2 Annual Incidence of Active Tuberculosisby Reactions to Two-Step Testing


Figure 3 Effect of age of BCG vaccination on two-step TST (PPD-T2) reactions amongCanadian-born health workers

Tuberculin conversion

Tuberculin conversion is defined as an increase in

tuberculin reactions of at least 6 mm following repeated

tuberculin testing, which is due to new mycobacterial

infection. Although easily stated, it can be very difficult to

distinguish tuberculin reactions due to conversions from

those due to boosting. The most important determinant is

the clinical situation. A positive tuberculin reaction, one

to four weeks after an initial negative test in the absence

of any known exposure to tuberculosis or other

mycobacteria is highly likely to be due to the booster

phenomenon. On the other hand, an individual who was

tuberculin negative and is now tuberculin positive eight

weeks following BCG vaccination, or exposure to a highly

contagious TB case is very likely to have demonstrated

conversion. Difficulties of interpretation occur when the

clinical situation falls between these two extremes.

The interval between primary infection and tuberculin

skin test conversion was carefully defined in two early

studies. In these studies, following a well-defined brief

exposure, 172 patients developed tuberculin conversion,

often with other manifestations of primary TB infection.

As shown in Figure 4, the great majority of documented

conversions occurred within six weeks and 100% within

eight weeks following the date of exposure. In addition, of

Table 3a Effect of BCG vaccination on two-step tuberculin testing

Table 3b Effect of BCG vaccination on 2-step tuberculin testing – Summary

Table 4 Effect of non-tuberculous mycobacteria (NTM) on two-step testing

Notes:1 Definitions of Booster : T1 < 10 mm, T2 ≥ 10 mm and increased by at least 6 mm. 2 TST: Tuberculin skin test – test material derived from M. tuberculosis – either RT-23, PPD-S, or PPD-T3 NTM: Non-tuberculous mycobacteria – test material derived from M. avium (51), M. intracellulare (17, 18, 52, 53), or M. scrofulaceum (18) Overall prevalence 1.3% in all tested. Data not shown separately but inferred that prevalence same in both groups.

Country (City) No Response to NTM3 Sensitivity to NTM3 (Ref)

Total (N) Initial TST2 Second TST2 Total (N) Initial TST2 Second TST1

(+) (%) (+) (%) (+) (%) (+) (%)

USA (Chicago) 110 1.3%3 1% 103 1.3%3 13% (18)

Canada (Montreal) 252 2.5% 1.6% 25 4% 12% (17)

Age vaccinated Number Positive TST2 (10+ mm) (Ref)

Studies (N) Subjects (N) Initial (%) Second (%)

0-1 7 1469 6.3% 9.9% (14, 17, 42-47)

5 and older 6 3159 43% 18% (15, 17, 43, 44, 47-50)

Author (Ref) Year Setting No. of Subjects Age Vaccinated Age Tested % with Booster1

Sepulveda (14) 1988 Chile 36 0-1 6 31%

Friedland (16) 1990 S. Africa 102 0-5 1-14 16%

Sepulveda (15) 1990 Chile 73 0-14 19-22 26%

Menzies (17) 1994 Montreal 380 0-1 17-25 8%

210 2-8 17-25 15%


200 BCG vaccine recipients, all developed tuberculin

reactions greater than 5 mm within six weeks (20). In all

experimental animals, positive tuberculin tests developed

two to three weeks after infection with M. tuberculosis

(21). Following inadvertent vaccination with M.

tuberculosis (the Lubeck disaster) tuberculin reactions

were positive in all children within seven weeks (22).

Figure 4 Interval From Primary Infection toTuberculin Skin Test Conversion(In 172 Persons with known timeof Infection)

The interval between acquisition of tuberculosis

infection and tuberculin conversion is important because

it determines the interval between the first and second

tuberculin test in contact investigations, this is often

called “window period”. Current recommendations are to

repeat the second test after 12 weeks. Given that all

conversions occurred within eight weeks, in all available

studies, it would seem prudent to perform the final

tuberculin test of a contact investigation eight weeks after

the end of exposure in order to detect new infections one

month sooner.

It is difficult to define the expected prevalence of

tuberculin conversions. In contact investigations, about

25% of all infected contacts will be identified from the

second tuberculin test. If the overall prevalence of

tuberculous infection in household contacts is 30-40%,

then the occurrence of tuberculin conversion should be

approximately one-quarter, or 7.5-10%. The prevalence

of conversion depends upon many other factors

determining the likelihood of transmission including

contagiousness of the index case and the duration and

environment of exposure.

The prognosis of tuberculin conversion is very

different from that of boosting. In cohorts with well-

documented tuberculin conversion, incidence has

ranged from 369 to 3,400 per 100,000. This difference

may be partially explained by different rates of boosting

(pseudo-conversion).

Distinguishing boosting from conversion

The most commonly utilized method to distinguish

these two reactions is size of reaction. The rationale for

this approach is that as the size of the repeated test result

increases, it is more likely due to conversion. Data from

Arkansas nursing home residents in Arkansas provide an

example of the different frequency distribution curves of

boosting (Figure 5) and conversion (Figure 6). In this

population, the boosting tuberculin reactions showed a

mode at reactions of 5-9 mm and larger reaction sizes

were progressively less frequent. On the other hand

tuberculin conversions were most frequently 15-19 mm

in size. In this population, a higher cut-point to define

conversion (i.e., 15 mm or greater) would improve

specificity, although with substantial loss of sensitivity.

Booster reactions showed a similar distribution in health

professional students and young workers in Montreal

but tuberculin conversions could not be so easily

distinguished from booster reactions.

Figure 5 Two Step TST (Booster Reactions)(Arkansas Nursing Home Residents–by Age)

Given the markedly different mechanisms—meaning

and risk of TB associated with these two reactions—it is

important to distinguish them as accurately as possible.

Table 4 summarizes the predictive value that an increase

in tuberculin reaction would represent true conversion in

different populations and clinical situations. These

predictive values show that household contacts with an

increased tuberculin reaction more likely have true

conversion than boosting and accordingly have high risk

of TB disease, regardless of the population group. On the

other hand, repeated testing of casual contacts is more

likely to result in increased tuberculin reactions due to


boosting. Health care workers are much more likely to

have boosting than conversion, unless initial two-step-

testing pre exposure is performed.

Based on Table 4, it can be recommended that casual

contacts should be tuberculin tested only once, 8-10

weeks after the exposure ends. This should not be

applied to high-risk casual contacts such as very young

children or HIV infected or patients with other immuno-

compromising conditions. This recommendation does

not apply to unusual circumstances such as, if there are

several secondary active cases among the contacts already

investigated, indicating a very highly contagious source

case or there was very prolonged exposure among the

casual contacts. However, as shown in Table 4, the vast

majority of casual contacts, if they are tested twice, will

be misdiagnosed as conversion when in fact they had

boosting. Performing a single test would reduce the

probability of such misdiagnosis, and the resultant

unnecessary (and potentially harmful) therapy for latent

TB infection.

Figure 6 Conversion tuberculin reactions(Arkansas Nursing Home Residents–by Age)

Notes:

* Likelihood of boosting from data in Table 2.

† Likelihood of conversion for second tuberculin test repeated 8 to 12 weeks post end of exposure (from ref. 19, 54)

‡ Conversion of 1% after 1 year of work based on ARI of 1% in clinical workers exposed to TB patients (55)

Table 5 Likelihood that a positive test from second sequential test represents conversion(Contacts and health care workers – 20-year-old adult)

Mycobacterial exposure Likelihood of: Positive predictive

value

BCG NTM M.TB Boosting* Conversion† For Conversion

(Given) (prevalence) (prevalence) (%) (%) (%)

Household Contact

Northern USA/Canada None 10% 1% 1.5% 18% 92%

Southern USA None 50% 1% 6.5% 18% 77%

Africa/Asia Infancy 50% 33% 23% 18% 49%

Western Europe Older 10% 1% 19.5% 18% 53%

Eastern Europe Older 10% 18% 24% 18% 48%

Casual Contact

Northern USA/Canada None 10% 1% 1.5% 4.5% 75%

Southern USA None 50% 1% 6.5% 4.5% 42%

Africa/Asia Infancy 50% 33% 23% 4.5% 17%

Western Europe Older 10% 1% 19.5% 4.5% 19%

Eastern Europe Older 10% 18% 24% 4.5% 16%

Health Care Worker

Northern USA/Canada None 10% 1% 1.5% 1%‡ 40%

Southern USA None 50% 1% 6.5% 1% 13%

Western Europe Older 10% 1% 19.5% 1% 5%


Tuberculin reversion

Serial tuberculin testing has also revealed that

tuberculin reversion may occur (23). In some

populations such as South African school children with

reactions of 14 mm or greater and children in Houston

treated for primary TB, reversion was seen in only 5%

after four years (24) or 8% after ten years (25). However

reversion is more common if initial tuberculin reactions

are only 5-9 mm (25) or if the test was positive because

of the booster phenomenon (26, 27). The phenomenon

of reversion emphasizes that once a tuberculin reaction

reaches 10 mm or more, further tuberculin testing

should not be done, because the results will be

uninterpretable. There are simply no data available upon

which to base the clinical management of an individual

whose tuberculin test was positive, then reverted to

negative but then later became positive again.

References

(1) Perez-Stable EJ, Slutkin G. A demonstration of lack

of variability among six tuberculin skin test readers.

Am J Public Health 1985; 75:1341-1343.

(2) Erdtmann FJ, Dixon KE, Liewellyn CH. Skin testing

for tuberculosis. Antigen and observer variability.

JAMA 1974; 228(4):479-481.

(3) Furcolow ML, Watson KA, Charron T, Lowe J. A

comparison of the tine and mono-vacc tests with the

intradermal tuberculin test. Am Rev Resp Dis 1967;

96:1009-1027.

(4) Thompson NJ, Glassroth JL, Snider DE, Farer LS.

The booster phenomenon in serial tuberculin

testing. Am Rev Resp Dis 1979; 119:587-597.

(5) Gordin FM, Perez-Stable EJ, Flaherty D, Reid ME,

Schecter G, Joe L et al. Evaluation of a third

sequential tuberculin skin test in a chronic care

population. Am Rev Respir Dis 1988; 137:153-157.

(6) Alvarez S, Karprzyk DR, Freundl M. Two-stage skin

testing for tuberculosis in a domiciliarly population.

Am Rev Respir Dis 1987; 136:1193-1196.

(7) Van den Brande P, Demedts M. Four-stage

tuberculin testing in elderly subjects induces age-

dependent progressive boosting. Chest 1992;

101:447-450.

(8) Barry MA, Regan AM, Kunches LM, Harris ME, et al.

Two-stage tuberculin testing with control antigens in

patients residing in two chronic disease hospitals.

JAGS 1987; 35:147-153.

(9) Morse DL, Hansen RE, Grabau JC, Cauthen G,

Redmond SR, Hyde RW. Tuberculin conversions in

Indochinese refugees. Am Rev Respir Dis 1985;

132:516-519.

(10) Cauthen GM, Snider DE, Onorato IM. Boosting of

tuberculin sensitivity among Southeast Asian refugees.

Am J Respir Crit Care Med 1994; 149:1597-1600.

(11) Menzies RI, Vissandjee B, Amyot D. Factors

associated with tuberculin reactivity among the

foreign-born in Montreal. Am Rev Resp Dis 1992;

146:752-756.

(12) Morse DL, Hansen RE, Swalbach G, Redmond

SR, Grabau JC. High rate of tuberculin conversion

in Indochinese refugees. JAMA 1982;

248(22):2983-2986.

(13) Veen J. Aspects of temporary specific anergy to

tuberculin in Vietnamese refugees. 1-119. 1992.

Amsterdam: K.N.C.V. Ref Type: Serial (Book,

Monograph).

(14) Sepulveda RL, Burr C, Ferrer X, Sorensen RU.

Booster effect of tuberculin testing in healthy 6-year-

old school children vaccinated with Bacillus

Calmette-Guerin at birth in Santiago, Chile. Pediatr

Infect Dis J 1988; 7(578):581.

(15) Sepulveda R, Ferrer X, Latrach C, Sorensen R. The

influence of Calmette-Guerin Bacillus immunization

on the booster effect of tuberculin testing in healthy

young adults. Am Rev Resp Dis 1990; 142:24-28.

(16) Escamilla B et al. Fungal content of dust and air

samples from asthamtic children’s homes in Mexico

City. Aerobiologia 1993.

(17) Menzies RI, Vissandjee B, Rocher I, St.Germain Y.

The booster effect in two-step tuberculin testing

among young adults in Montreal. Ann Intern Med

1994; 120:190-198.

(18) Richards NM, Nelson KE, Batt MD, Hackbarth D,

Heidenreich JG. Tuberculin test conversion during

repeated skin testing, associated with sensitivity to

nontuberculous mycobacteria. Am Rev Resp Dis

1979; 120:59-65.

(19) Ferebee SH. Controlled chemoprophylaxis trials in

tuberculosis. Adv Tuberc Res 1969; 17:28-106.

(20) Guld J. Response to BCG vaccination. In: Palmer

CE, Magnus K, Edwards LB, editors. Studies by The

WHO Tuberculosis Research Office. Geneva: World

Health Organization, 1953: 51-56.

(21) Youmans GP. Tuberculosis. Philadelphia: W.B.

Saunders Company, 1979.


(22) Triep WA. De tuberculinereactie. (Dutch only).

1957. Royal Netherlands Tuberculosis Association

(KNCV), The Haag, Netherlands. Ref Type: Serial

(Book, Monograph).

(23) Dahlstrom AW. The instability of the tuberculin

reaction. Observations on dispensary patients, with

special reference to the existence of demonstrable

tuberculous lesions and the degree of exposure to

tubercle bacilli. Am Rev Tuberc 1940; 42:471-487.

(24) Felten MK, Van Der Merwe CA. Radom variation in

tuberculin sensitivity in schoolchildren. Am Rev

Resp Dis 1989; 140:1001-1006.

(25) Hsu KHK. Tuberculin reaction in children treated

with Isoniazid. Am J Dis Child 1983; 137:1090-

1092.

(26) Gordon FM, Perez-Stable EJ, Reid M, Schecter G,

Cosgriff L, Flaherty D et al. Stability of positive

tuberculin tests: Are boosted reactions valid? Am Rev

Respir Dis 1991; 144:560-563.

(27) Perez-Stable EJ, Flaherty D, Schecter G, Slutkin G,

Hopewell PC. Conversion and reversion of

tuberculin reactions in nursing home residents. Am

Rev Resp Dis 1988; 137:801-804.

(28) Chaparas SD, Mac Vandiviere H, Melvin I, Koch G,

Becker C. Tuberculin Test: Variability with the

Mantoux procedure. Am Rev Resp Dis 1985;

132:175-177.

(29) Bearman JE, Kleinman H, Glyer VV, LaCroix OM. A

Study of Variability in Tuberculin Test Reading. Am

Rev Resp Dis 1964; 90:913-918.

(30) Loudon RG, Lawson JR, Brown J. Variation in

tuberculin test reading. Am Rev Respir Dis 1963;

87:852-861.

(31) Fine MH, Furculow ML, Chick EW, Bauman DS,

Arik M. Tuberculin skin test reactions. Am Rev Respir

Dis 1972; 106:752-758.

(32) Howard TP, Solomon DA. Reading the tuberculin

skin test: who, when and how? Arch Intern Med

1988; 148:2457-2459.

(33) Pouchot J, Grasland A, Collet C, Coste J, Esdaile JM,

Vinceneux P. Reliability of tuberculin skin test

measurement. Ann Intern Med 1997; 126(3):210-214.

(34) Bass JB, Serio RA. The use of repeated skin tests to

eliminate the booster phenomenon in serial tuberculin

testing. Am Rev Resp Dis 1981; 123:394-396.

(35) Gross TP, Israel E, Powers P, Cauthen G, Rose J. Low

prevalence of the booster phenomenon in nursing-

home employees in Maryland. Maryland Med J

1984; 35:107-109.

(36) Slutkin G, Perez-Stable EJ, Hopewell PC. Time

course and boosting of tuberculin reactions in

nursing home residents. Am Rev Resp Dis 1986;

134:1048-1051.

(37) Burstin SJ, Muspratt JA, Rossing TH, et al. Studies of

the dynamics of reactivity to tuberculin and

Candida antigen in institutionalized patients. Am

Rev Respir Dis 1986; 134:1072-1074.

(38) Webster CT, Gordin FM, Matts JP, Korvick JA,

Miller C, et al. Two-stage tuberculin skin testing in

individuals with human immunodificiency virus

infection. Am J Respir Crit Care Med 1995;

151:805-808.

(39) Hecker MT, Johnson JL, Whalen CC, Nyole S,

Mugerwa RD, Ellner JJ. Two-step tuberculin skin

testing in HIV-infected persons in Uganda. Am J

Respir Crit Care Med 1997; 155:81-86.

(40) Lifson AR, Grant SM, Lorvick J, Pinto FD, He H,

Thompson S et al. Two-step tuberculin skin testing

of injection drug users recruited from community-

based settings. J Tuberc Lung Dis 1997; 1(2):128-134.

(41) Groth-Peterson E, Knudsen J, Wilbeck E.

Epidemiological basis of tuberculous eradication in

Denmark. Bull Wld Hlth Org 1959; 21(1):5-49.

(42) Lifschitz M. The value of the tuberculin skin test as

a screening test for tuberculosis among BCG-

vaccinated children. Pediatrics 1965; 36:624-627.

(43) Margus JH, Khassis Y. The tuberculin sensitivity in

BCG vaccinated infants and children in Israel. Acta

Tuberc Pneumonol Scand 1965; 46:113-122.

(44) Joncas JH, Robitaille R, Gauthier T. Interpretation of

the PPD skin test in BCG-vaccinated children. Can

Med Assoc J 1975; 113:127-128.

(45) Karalliede S, Katugha LP, Uragoda CG. The

tuberculin response of Sri Lankan children after

BCG vaccination at birth. Tuberc 1987; 68:33-38.

(46) American Conference of Governmental Industrial

Hygenists. Guidelines for assessment and sampling

of saprophytic bioaerosols in the indoor

environment. Applied Industrial Hygiene 1987;

2:R10-R16.

(47) Menzies RI, Vissandjee B. Effect of Bacille Calmette-

Guerin vaccination on tuberculin reactivity. Am Rev

Resp Dis 1992; 145:621-625.

(48) Comstock GW, Edwards LB, Nabangwang H.

Tuberculin sensitivity eight to fifteen years after BCG

vaccination. Am Rev Resp Dis 1971; 103:572-575.


(49) Bahr GM, Chugh TD, Behbehani K, Shaaban MA, et

al. Unexpected findings amongst the skin test

responses to mycobacteria of BCG vaccinated

Kuwaiti school children. Tuberc 1968; 68:105-112.

(50) Horwitz O, Bunch-Christensen K. Correlation

between tuberculin sensitivity after 2 months and 5

years among BCG vaccinated subjects. Bull Wld Hlth

Org 1972; 47:49-58.

(51) Lind A, Larsson O, Bentzon MW, Magnusson M,

Olofson J, Sjogren I et al. Sensitivity to sensitins

and tuberculin in Swedish children. Tuberc 1991;

72:29-36.

(52) Jeanes CWL, Davies JW, McKinnon NE. Sensitivity

to “Atypical” Acid-Fast Mycobacteria in Canada.

CMAJ 1969; 100:1-8.

(53) Menzies RI. The determinants of the prevalence of

tuberculous infection among Montreal

schoolchildren, Thesis for M.Sc. in Epidemiology

and Biostatistics. Montreal: McGill University, 1989.

(54) Zaki MH, Lyons HA, Robins AB, Brown EP.

Tuberculin sensitivity. New York State Journal of

Medicine 1976; 76:2138-2143.

(55) Menzies RI, Fanning A, Yuan L, FitzGerald J.M.

Hospital ventilation and risk of tuberculous

infection in Canadian Health Care Workers. Ann

Intern Med 2000; 133(10):779-789.


TUBERCULIN SENSITIVITY PRODUCED BYMYCOBACTERIA OTHER THAN THEMYCOBACTERIUM TUBERCULOSIS COMPLEX

George Comstock, M.D.,

Dr.P.H. (deceased)

Dr. Comstock was Professor

Emeritus in the Department of

Epidemiology of the Bloomberg

School of Public Health at Johns

Hopkins University, Baltimore,

Maryland

“Nontuberculous mycobacteria” is a name

suggested by Wolinsky as the “least

offensive” term for this numerous and diverse group of

mycobacteria. The name itself indicates that the group is

defined by exclusion (1). Primarily excluded are members

of the Mycobacterium tuberculosis complex. This group is

also referred to as “tubercle bacilli” and consists of three

human pathogens and two nonpathogens, which are also

used as vaccines (2). The pathogenic members are

Mycobacterium tuberculosis, Mycobacterium bovis and

Mycobacterium africanum; the two nonpathogens are

Mycobacterium bovis BCG and Mycobacterium microti, the

vole bacillus.

M. bovis is believed to cause tuberculin reactions that

are somewhat larger on average than those due to M.

tuberculosis (3). M. bovis BCG causes highly variable

degrees of tuberculin sensitivity depending on the

particular strain of vaccine. Post-vaccinal sensitivity

ranges from relatively few and weak reactions to reactions

similar in size to those caused by M. bovis (4,5). M.

leprae is also excluded from most discussions of

nontuberculous mycobacteria.

During the first half of the last century,

nontuberculous mycobacteria were hardly ever

mentioned or considered. By century’s end, they had

become well known, largely because of the widespread

use of diagnostic cultures for M. tuberculosis and because

of the disease they caused among persons whose immune

systems had been compromised by the human

immunodeficiency virus.

In the 1930’s, one of the dogmas of tuberculosis was

that a positive reaction to any tuberculin test was

pathognomonic of tuberculosis infection (6,7). Even

then, however, this dogma was being challenged, most

forcibly by Leslie L. Lumsden, the U. S. Public Health

Service’s foremost shoe-leather epidemiologist (8).

Lumsden was studying tuberculosis in two Southern

counties—Giles County, Tennessee with a high

tuberculosis death rate, and Coffee County, Alabama,

with a low rate (9). His team started their studies by

tuberculin testing school children to obtain an index of

the prevalence of infection with Mycobacterium

tuberculosis. They used a commercially available PPD-

tuberculin produced in tablet form, presumably made by

Parke-Davis according to Dr. Seibert’s procedures. The

dosage selected was 0.0005 mg in 0.1 ml of diluent, one-

tenth the usual second strength, and in theory, roughly

equivalent to 25 Tuberculin Units (TU). When the

proportion of positive reactors in Giles County was found

to be lower than expected, a series of comparisons with

tuberculin preparations used in other field studies was

started, all in the same doses of one-tenth of the standard

second-strength. Each of the four other preparations was

compared with the tuberculin used originally. The

characteristics of the four groups of children were not

stated except that all were in school in Coffee County.

The results of the comparisons in terms of the

proportions of positive reactors are shown in figure 1. The

tuberculin to which the others were compared was the

tablet preparation, labeled PPD-1 in this figure. It is clear

that the prevalence of positive reactors in these school

children varied markedly, depending on which tuberculin

preparation was being used. Lumsden concluded that

“skin testing with any of the tuberculin preparations now

on the market or otherwise amply available is of

questionable value or definitely futile” (9).

Because Lumsden had usually been right even when

his conclusions were controversial, it was decided to

repeat his work under circumstances that would insure

that all technical aspects of administering and reading the

tuberculin tests were beyond question (10-12). The site


of the new study was Washington County, Maryland,

where tuberculosis studies were already in progress

under the direction of Carroll Palmer, then Director of

Research for the Bureau of Child Hygiene. The

tuberculin preparations were the same as those in the

Lumsden study, although most comparisons were against

a new standard (presumably from Florence Seibert). The

usual first and second doses of 1 and 250 TU were

administered, the second dose being given only to

persons with negative reactions to the first dose.

Figure 1 Comparisons of percent positive reactorsto a standard tuberculin (PPD-1) andone of four other tuberculins (PPD-3, OT M, OT 511 or OT 711) among schoolchildren in Coffee County, Alabama.

Source: Lumsden, et al (1939)

Figure 2 shows the proportions of positive reactors to

six different pairs of tuberculin preparations. The lower

crosshatched portions of the bars indicate the

proportions reacting to the first weaker dose of

tuberculin. Note the similarity among these proportions,

especially among the school children. The total heights

of the bars show the proportions reacting to either the

first or second dose. The considerable variations in

positive reactors to the different preparations are almost

entirely due to variations in reactions to the second dose.

The assembled experts agreed that all technical

aspects of tuberculin testing had been done satisfactorily.

They also agreed that the second stronger dose of

different tuberculins caused different proportions of

reactions. Only two persons appear to have come away

from the meeting with any hypothesis as to the cause of

the differences. Esmond Long, director of the Phipps

Institute in Philadelphia, and Carroll Palmer, a young

medical statistician, both felt that there had to be more

than one cause of tuberculin sensitivity. Looking for

those additional causes was to become a major part of

Palmer’s subsequent career.

Figure 2 Comparisons of percent positive reactorsto different pairs of tuberculins amongschool children and adults inWashington County, Maryland, in 1938.

Source: Unpublished data.

FINDING MORE THAN ONE CAUSE OF

TUBERCULIN SENSITIVITY

Ten years later, Palmer had his big chance. The

International Tuberculosis Campaign was organized in

1948 to combat the tuberculosis epidemic facing many

nations in the wake of World War II (13). Its major

activity was mass BCG vaccination, preceded by

tuberculin testing which was then considered necessary

to identify persons eligible for vaccination. Millions of

children were tuberculin tested in the next three years.

Palmer, as head of the associated Tuberculosis Research

Office, was able to obtain reliable tuberculin test results

from many of these countries by including in the

Campaign special teams of highly trained nurses. Many

practical and administrative lessons were learned during

the course of the Campaign. For subsequent research,

the most important lesson was that the mean and

standard deviation were far more useful descriptors of

tuberculin reactions among groups than merely reporting

reactions as positive or negative, or even than the more

sophisticated classification into five groups—negative,

and four degrees of positivity. To obtain mean reaction


sizes required that the diameters of induration be

measured carefully in millimeters.

In Figure 3, the histograms represent the distributions

of reactions sizes to 10 TU of PPD-tuberculin among

school children in 12 different countries, while the line

graphs represent the distributions of reaction sizes

among tuberculosis hospital patients in the same

countries (14). For tuberculosis patients, the

distributions are unimodal and tend to center at 15-16

mm, regardless of country of residence. In fact, they are

remarkably uniform, considering all the problems

involved in trying to obtain uniformity in tuberculin

testing. For school children, there are marked variations

in the shapes of the histograms. Most of these

distributions are bimodal. The purest is that for

Northern U.S.A. (Michigan) with hardly any induration

except for a small group with the same range of

induration as tuberculosis patients. In other places, the

“valley” between the left and right-hand modes is “filled

in” to various degrees until one comes to distributions

like those in South India and the southern U.S. (Georgia)

where no right-hand mode is distinguishable. It

appeared to Palmer and his colleagues that there had to

be some agents other than tuberculosis that were causing

varying degrees of tuberculin sensitivity. Lacking specific

candidates for the causal agents, they termed this

nontuberculous tuberculin sensitivity “nonspecific”.

This term has since been used for tuberculin sensitivity

caused either by nontuberculous mycobacteria or by the

two nonpathogenic members of the M. tuberculosis

complex, M. bovis BCG or M. microti.

Palmer’s extensive studies of tuberculin sensitivity and

tuberculosis among some 22,000-student nurses had also

confirmed his belief in nonspecific tuberculin sensitivity

(15). As shown in Figure 4, increasing degrees of contact

with tuberculosis increased the likelihood of reacting to

the 5 TU dose of tuberculin, but beyond that, degree of

contact was not related to the intensity or size of the

reaction (15). Reactions only to the second dose bore no

relationship to degree of contact. These results led to the

conclusion that nonspecific sensitivity was caused by a

nontuberculous organism with a different mode of

transmission. Such an organism or organisms had to be

“antigenically related to the tubercle bacillus, highly

prevalent in certain geographic areas, and apparently

non-pathogenic for human beings” (15).

Figure 3 Distributions of reaction sizes to 5tuberculin units of PPD-tuberculinamong school children (histograms) and tuberculosis patients (heavy lines) in various countries.

Source: Reproduced with permission from WHO TB Research Office(1955).

Some members of the group of nontuberculous

mycobacteria were logical candidates. Nontuberculous

mycobacteria had been recognized as early as 1885 and

had occasionally been found to be associated with cases of

disease (1). In 1956, the group from Battey State Hospital

in Georgia reported a large case series consisting of 64

patients who had a disease similar to tuberculosis except

that the mycobacterium repeatedly isolated from sputum

specimens produced smooth buff-colored colonies, none

of which were virulent for guinea pigs, and were thus

clearly not Mycobacterium tuberculosis (16). Palmer was

able to obtain 87 isolates of this Battey bacillus for animal

experiments. He was also able to obtain the comparative

tuberculin tests with PPD-S and a PPD made from the so-

called Battey bacillus (PPD-B) for 84 patients with disease

produced by the Battey bacillus and for 1,434 patients

with typical M. tuberculosis in their sputa (17).

��

��

��

��

��

��

��

��

�

� � �


Figure 4 Distributions of reaction sizes to 5 and250 tuberculin units of PPD-tuberculinamong student nurses by history ofcontact with tuberculosis.

Source: Reproduced with permission from Palmer (1953)

The results of duplicate tests with PPD-S and PPD-B

among guinea pigs infected with the H37RV strain of M.

tuberculosis, the Battey bacillus, or with both are shown in

Figure 5 (17). Among animals infected with only one

organism, the homologous antigen caused the largest

reaction in almost every instance. Among animals

infected with both organisms, sensitivity to PPD-S

predominated, the pattern in the correlation diagram

being nearly the same as the pattern of animals infected

only with M. tuberculosis.

Results of dual testing with PPD-S and PPD-B among

patients at Battey State Hospital were remarkably similar

to those among guinea pigs (Table 1) (17). Human

studies can never be as clear-cut as those among

experimental animals where the sources of infection can

be known with certainty. Although we can be sure that

patients with sputum positive for M. tuberculosis or the

Battey bacillus (now known as M. avium/intracellulare)

are infected with those organisms, we cannot tell which

patients had been infected with more than one

mycobacterium. Even with this uncertainty, the results of

dual testing with PPD-S and PPD-B showed that

tuberculosis patients are likely to have larger reactions to

the homologous antigen, PPD-S,

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

��

Figure 5 Correlations of reaction sizes to 0.0005mg PPD-S and PPD-B among guinea pigsinfected with M. tuberculosis (H37Rv),the Battey bacillus or both.

Source: Reproduced with permission from Palmer, et al (1959)

than to PPD-B, and patients infected with Battey bacillus

are more likely to have larger reactions to PPD-B.

Returning to findings among guinea pigs, Palmer used

the known patterns of sensitivity to PPD-S caused by both

the specific infection with M. tuberculosis and the

nonspecific infection by the Battey organism to show what

would be expected with varying mixtures of these two

infections (17). The histogram that would result with 5%

of the animals infected with M. tuberculosis and 25% with

the Battey bacillus is shown in Figure 6. All guinea pigs

with reactions to PPD-S of 10 mm or larger had been

infected with tubercle bacilli, those with reactions

between 5 and 10 mm were a mixture of tuberculosis- and

Battey-infected animals, and those with less than 6 mm of

induration were mostly uninfected, with a few Battey-

infected animals mixed in. These results, later duplicated

with other nontuberculous mycobacteria, showed that the

larger the reaction to the intermediate dose of tuberculin,

5 or 10 TU, the more likely that the cause was an infection

with M. tuberculosis. Conversely, small reactions to the

intermediate dose, and all reactions only to the strong

second dose, were highly likely to be caused by some

nontuberculous mycobacteria.


Table 1 Results of Duplicate Tests with PPD-S andPPD-B among Patients with SputumPositive for M. tuberculosis or BatteyBacillus

This particular guinea pig experiment was

approximated among humans by conducting dual

testing with PPD-S and PPD-B and assuming that the

larger of the two reactions indicated the cause of the

sensitivity. Figure 7 shows the results of dual testing in

Minnesota and India (17). The solid black portions of

the bars represent persons whose reactions to PPD-S

were larger than those to PPD-B, and who are presumed

to be those infected with M. tuberculosis. The striped

portions of the bars represent persons whose reactions

to PPD-B are equal to or larger than those to PPD-S, and

who are believed to have been infected with an

organism antigenically similar to the Battey bacillus.

Again, the similarity of these two distributions to the

previous one for guinea pigs is striking and consistent

with the belief that larger reactions indicate the cause of

the tuberculin sensitivity.

LONGITUDINAL STUDIES CAN HELP ASSESS

TUBERCULIN SENSITIVITY

Another way to assess the significance of varying

degrees of tuberculin sensitivity is through longitudinal

studies. If smaller tuberculin reactions are largely due to

nontuberculous mycobacteria, subsequent tuberculosis

incidence should be lower among persons with weak

tuberculin sensitivity. An early illustration that this is true

was found in a controlled trial of BCG vaccination among

school children in Muscogee County, Georgia (18). In

1947, participants in this trial were tested with RT-18, a

purified tuberculin produced by the State Serum Institute

in Copenhagen, Denmark. Children who did not react to

the 5 TU dose were given the 100 TU dose. The incidence

of tuberculosis among the nonvaccinated children after 12

years of follow-up is shown in Table 2. The average annual

incidence of tuberculosis was much greater among those

who reacted to the 5 TU dose. Incidence among those

Tuberculous Battey BacillusPatients Patients

No. % No. %

S > B+1 1,234 86.0 18 21.4

S = ± 1 150 10.5 13 15.5

S < B-1 50 3.5 53 63.1

Total 1,434 100.0 84 100.0

Figure 6 Expected frequency distribution ofreaction sizes to 0.0005 mg PPD-S inguinea pigs if 5% were infected with M.tuberculosis (dark shading), 25% withthe Battey bacillus (striped shading) and70% with nothing (stippled shading).


who reacted only to the 100 TU dose was not significantly

different than among nonreactors.

Additional evidence comes from the USPHS study of

Navy recruits (19). The results of a five-year follow-up

study are shown in Table 3. The recruits were initially

tested with 5 TU of PPD-S and an equivalent dose of

either PPD-B or PPD-G, a PPD prepared from M.

scrofulaceum. Among recruits with less than 6 mm of

induration to PPD-S, size of reaction relative to the other

antigens had little relationship to subsequent tuberculosis

incidence. Among those with reaction diameters to PPD-

S of 6-17 mm, relative size of induration to PPD-B or G

was important. Those with PPD-S reactions smaller than

those to PPD-B or PPD-G had a very low rate of

subsequent tuberculosis; those with larger PPD-S

reactions had a subsequent risk 7.5 times greater,

essentially the same as persons with PPD-S reactions of 18

mm or more, persons whom we have already seen were

virtually all infected with M. tuberculosis.

THE CHALLENGE OF DIFFERENTIATING

NONSPECIFIC REACTIONS

Finally there is the matter of how to estimate the

proportion of reactors to 5 TU of PPD-tuberculin that are

due to tuberculous and nontuberculous infections. Two

assumptions are necessary, both backed up by extensive

animal experiments and human observations. The first is

that the distributions of reactions caused by tuberculosis

are symmetrical and second, that they have a single mode

at approximately 16-17 mm. Applying these two

assumptions to a distribution of reaction sizes allows the


creation of a distribution of tuberculous reactions by

producing a mirror image of reactions greater than 16-17

mm and joining that mirror image to its original to

produce a theoretical symmetrical distribution of specific

tuberculous reactions.

Figure 7 Frequency distributions of reaction sizes to 5 TU of PPD-S and othermycobacterial antigens in generalpopulation groups in Minnesota (adults) and India (all ages).

Note: Dark shading is for persons whose reaction sizes to PPD-S werelarger than to the other mycobacterial antigen; striped shading is forpersons whose reaction sizes to PPD-S were the same or smaller thanto the other antigen; and stippled shading is for persons whosereaction sizes to PPD-S were less than 6 mm regardless of reaction sizesto other antigens.


Table 2 Tuberculosis Incidence among Muscogee County, Georgia schoolchildren, 1947-1959, by initial reaction to two doses of tuberculin RT-18 (18)

* p difference - 0.44

Tuberculous CasesRate/

N No. 100,000/yr

Reactors, 5 T.U. 1,482 24 135.0

Reactors, 100 T.U. only 3,768 5 11.1*

Non reactors, 100 T.U. 2,341 2 7.1*

Table 3 Tuberculosis Incidence among U. S. Navypersonnel in a five-year period afterenlistment by initial reactions to 5 T.U.PPD-S and Similar Doses of PPD-B orPPD-G (19)

This has been done in Figure 8, which represents the

various reaction sizes to 5 TU of RT 19-20-21 among

junior and senior high school students in Muscogee

County, Georgia, and its neighbor, Russell County,

Alabama (20). Because only a single antigen was used in

the testing, it is not possible to tell which individual

students with reactions between 9 and 16 mm are

infected with tubercle bacilli. However, the mirror image

method gives the estimated distribution of tuberculous

reactions. We can now see that if children with 10 or

more mm are called positive, that definition will include

a large number who have not been infected with tubercle

bacilli. Using the current definition of positive—namely

15 or more mm of induration—only a small proportion

would be misclassified as positive, but an appreciable

number of true positives would be called negative.

Obviously, this method of estimating the proportion of

tuberculous-infected persons is not perfect, but it does

allow an administrator to make reasonable estimates

based on the particular tuberculin being used and the

tendencies of the local test readers to exaggerate or

minimize diameters of induration.

A similar situation is portrayed in Figure 9. Again,

junior and senior high school students were the persons

tested (21). They lived in Washington County, Maryland,

an area with a much smaller proportion of nonspecific

reactors. But as in the previous figure, an important

feature was that the tests were administered and read by

a highly trained team of nurses, supervised in this

instance by Lydia Edwards. As before, a distribution of

reactions presumed to be tuberculous in origin has been

created by projecting to the left of the mode of 17 mm

the mirror image of the observed distribution to the right

Tuberculous CasesRate/

PPD-S PPD-B/G n Cases 100,000/yr

0 - 5 Any 1,058,122 384 7.3

6 - 17 > PPD-S 19,333 10 10.3

± PPD ± 1 11,934 22 36.9

< PPD-S 22,314 84 75.3

18+ Any 13,180 49 74.4


of that mode. The resulting distribution of reactions

estimated to be due to infections with tubercle bacilli is

shown by cross-hatching. As in Muscogee and Russell

Counties, classifying reactions of 10 or more mm as

positive includes almost all the tuberculous reactions but

at the cost of also including a fair proportion of

nonspecific reactions. Increasing the definition of

positive to 15 or more mm produces a group almost all

of whom are tuberculous infected but misses nearly half

of the truly infected. But as before, an administrator in

this situation has information to allow an informed

decision of what the cut point between positive and

negative should be and can see what the consequences

would be of applying national criteria locally.

Figure 8 Frequency distribution of reaction sizesto 5 tuberculin units of RT 19-20-21among school children in Georgia andAlabama (with reaction sizes estimatedto be due to infection with M. tuberculosis shown in darker shading).

Source: Comstock (1960)

SMOOTHING DISTRIBUTION

Not all health departments have tuberculin testers and

readers who can produce such smooth distributions as

those in Figures 8 and 9. Figure 10 shows distributions

of reaction sizes from two studies conducted with local

personnel who claimed to have considerable experience

in tuberculin testing. The study conducted in the 1970’s

produced an extreme example of an all too familiar

phenomenon, terminal digit preference. This is the

Figure 9 Frequency distribution of reaction sizes

to 5 tuberculin units of PPD-S among

school children in Washington County,

Maryland, with reaction sizes estimated

to be due to infection with M. tuberculosis

shown by cross-hatching.

Source: Kuemmerer and Comstock (1960)

tendency, when measuring entities that have indistinct

boundaries, to record an excess of measurements ending

in particular digits, usually 5’s or 0’s. In this instance, the

tendency was so extreme that no method of smoothing

the distribution is feasible and applying the mirror image

method of estimating the proportion of tuberculous

infected is out of the question.

In the study done in the 1990’s, the situation was

much better. An attempt to smooth the distribution by

applying a three-point moving average is shown by the

solid line. While an attempt at smoothing seemed

reasonable, it gave no indication of a single right-hand

mode. Furthermore, the excess of very large reaction sizes

in both studies exceeds anything seen elsewhere. Again,

terminal digit preference, though less extreme in 1990

than in the 1970 study, is still too prominent to allow

reasonable estimates by the mirror image method.


Figure 10 Frequency distributions of reactionsizes to 5 tuberculin units of PPD-tuberculin among persons tested in1970 and 1990 by staff of an unnamedhealth department.

Note: Heavy line is a 3-point moving average

There are two ways to minimize terminal digit

preference and to produce a relatively smooth

distribution. Because terminal digit preference is only

likely to occur when there is some degree of uncertainty

in measurement and therefore room for subjective

influences to occur, any way of insuring clear and definite

measurement points should go far toward solving the

problem. One method is to use the ballpoint pen method

of Sokol in which the pen is pressed along the skin toward

the area of induration from above and below until

resistance is felt (22). The distance between the ends of

the two lines is the sagittal diameter of induration, and

should be measurable with very little uncertainty.

Another method is to rely on visualization rather than

palpation, either with the finger or indirectly with a pen,

and to use a measuring device in which the scale is not

seen until the measurement has been made.

Visualization is important since what can be seen, can be

measured. Induration, even that which is difficult to

palpate, can easily be seen if diffuse light is allowed to

shine tangentially across the test site. Once the borders

of induration are seen, the transverse diameter should be

measured as mechanically as possible, paying no

attention to what the numbers may mean. Once the

diameter is recorded, it is time to become a health care

provider again and think what the reaction portends for

the care of the patient.

Some sort of a gauge is almost essential. A Dr.

Turnbull at a Southern Trudeau Society meeting about

1950 first suggested a simple, readily available and

inexpensive gauge. The “Turnbull gauge,” as we called it,

was merely a sewing gauge. At that time, a millimeter

scale had to be glued over the original scale, which was

in inches. The sliding pointer was then reversed so that

the knob that moved the pointer was on the side opposite

the scale, making it impossible to tell what the reading

was until after the gauge had been set. Then it could be

turned over and the reading recorded to the nearest mm

unit below the pointer, thereby avoiding fractions of a

millimeter and virtually all uncertainty. Today, sewing

gauges come with millimeter scales, making them usable

with only the simple reversal of the sliding pointer.

Figure 11 illustrates that the routine use of a modified

sewing gauge to measure tuberculin reactions can

minimize terminal digit preference. This is the

distribution of tuberculin reaction sizes recorded for the

790 persons tuberculin tested by personnel of the

Washington County Health Department in the year 2000

using the modified sewing gauge. The bars of the

histogram show the observed distribution. There is a

little evidence of preference for numbers ending in 5’s or

0’s but so little that the distribution is easily smoothed by

a 3-point moving average, shown here by the solid line.

Using the mirror image technique, the estimated

proportion of persons infected with M. tuberculosis is

obtained and is shown by the crosshatched part of the

smoothed distribution. It appears that a reasonable cut-

point between positive and negative reactions in this

population might be 12 mm of induration. To the left of

that line, most of the reactions can safely be assumed to

be nontuberculous; to the right, nearly all are assumed to

be tuberculous.

In summary, it is now clear that wherever

nontuberculous mycobacterial infections are found, there

will also be nonspecific sensitivity to the intermediate

doses of PPD-tuberculin. This is true for most of the

world except for high altitudes and high latitudes. If

tuberculin tests can be carefully and accurately

administered and if the ensuing reactions are measured

in ways that are objective and associated with minimal

uncertainty, analysis of the results can give reasonable


estimates of the proportions of reactions due to

tuberculous and to nontuberculous infections. In

general, the larger the reaction to 5 TU of PPD-

tuberculin, the more likely that it was caused by infection

with tubercle bacilli. Reactions with induration

measuring 17 mm or more in diameter are almost certain

to be due to infection with M. tuberculosis, M. bovis or

some strain of BCG. Reactions less than 10 mm are

almost all due to infections with nontuberculous

mycobacteria or most strains of BCG. For maximal

usefulness, the tuberculin test needs to be administered

with care and the resulting reactions must be read

without bias and with minimal terminal digit preference.

.Figure 11 Frequency distribution of reaction sizes

to 5 tuberculin units of PPD-tuberculinamong persons tested in 2000 by staffof Washington County (Maryland)Health Department.

Note: Heavy line is a 3-point moving average. Reactions estimated tobe due to infection with M. tuberculosis shown by cross-hatching.

Source: Data furnished by Mark Jameson, Washington County HealthDepartment.

References

(1) Wolinsky E. State of the art. Nontuberculous

mycobacteria and associated diseases. Am Rev

Respir Dis 1979; 119:107-159.

(2) Grosset J, Truffot-Pernot C, Cambau E. Bacteriology

of tuberculosis. In Reichman LB, Hershfield ES,

eds. Tuberculosis. A Comprehensive International

Approach (Second ed.). New York: Marcel Dekker,

2000, pp. 157-185.

(3) Magnus K. Epidemiologic basis of tuberculosis

eradication. 6. Tuberculin sensitivity after human

and bovine infection. Bull Wld Hlth Org 1967;

36:719-731.

(4) Shaw LW. Field studies on immunization against

tuberculosis. I. Tuberculin allergy following BCG

vaccination of school children in Muscogee County,

Georgia. Public Health Rep 1951; 66:1415-1426.

(5) Edwards LB, Palmer CE, Magnus K. BCG

vaccination. Studies by the WHO Tuberculosis

Research Office, Copenhagen. Chapter 3, Response

to BCG Vaccination. World Health Organization

Monograph Series, No. 12. Geneva: World Health

Organization, 1953, pp. 51-64.

(6) Griffith JPC, Mitchell AG. The Diseases of Infants and

Children (Second ed.). Philadelphia: W. S. Saunders

Company, 1938, p. 396.

(7) Committee on Diagnostic Standards, National

Tuberculosis Association. Diagnostic Standards and

Classification of Tuberculosis (1940 ed.). New York:

National Tuberculosis Association, 1940, p. 24.

(8) Furman B. A Profile of the United States Public

Health Service, 1798-1948. DHEW Publication No.

(NIH) 73-369. Washington D.C.: National Library

of Medicine, U. S. Department of Health, Education

and Welfare, 1973, pp. 290-292.

(9) Lumsden LL, Dearing WP, Brown RA. Questionable

value of skin testing as a means of establishing an

epidemiological index of tuberculous infection. Am

J Public Health 1939; 29:25-33.

(10) McKneely TB. An evaluation of the tuberculin test

as a screen in school case-finding programs. Trans

34th Annual Meeting of the National Tuberculosis

Association, 1938, pp. 290-296.

(11) Anon. Minutes of Hagerstown Tuberculosis

Conference, September 26-October 1, 1938.

Evening Session—September 30, 1938.

(Unpublished).


(12) Comstock GW. The Hagerstown Tuberculosis

Conference of 1938: A retrospective opinion

(editorial). Am Rev Respir Dis 1969; 99:119-120.

(13) Comstock GW. The International Tuberculosis

Campaign: A pioneering venture in mass

vaccination and research. Clin Infect Dis 1994;

19:528-540.

(14) WHO Tuberculosis Research Office. Further studies

of geographic variation in naturally acquired

tuberculin sensitivity. Bull Wld Hlth Org 1955;

12:63-83.

(15) Palmer CE. Tuberculin sensitivity and contact with

tuberculosis. Further evidence of nonspecific

sensitivity. Am Rev Tuberc 1953; 68:678-694.

(16) Crow HE, King CT, Smith CE, Corpe RF, Stergus I.

A limited clinical, pathologic and epidemiologic

study of patients with pulmonary lesions associated

with atypical acid-fast bacilli in the sputum. Am Rev

Tuberc Pulm Dis 1957; 75:199-222.

(17) Palmer CE, Edwards LB, Hopwood L, Edwards PQ.

Experimental and epidemiologic basis for the

interpretation of tuberculin sensitivity. J Pediatr

1959; 55:413-429.

(18) Comstock GW, Shaw LW. Tuberculosis studies in

Muscogee County, Ga. Controlled trial of BCG

vaccination in a school population. Public Health

Rep 1960; 75:583-594.

(19) Edwards LB, Acquaviva FA, Livesay VT.

Identification of tuberculous infected. Dual tests

and density of reaction. Am Rev Respir Dis 1973;

108:1334-1339.

(20) Comstock GW. A comparison of purified

tuberculins in the southeastern USA. Bull Wld Hlth

Org 1960; 23:683-688.

(21) Kuemmerer JM, Comstock GW. Sociologic

concomitants of tuberculin sensitivity. Am Rev Respir

Dis 1967; 96:885-892.

(22) Sokol JE. Measurement of delayed skin test

responses. N Engl J Med 1975; 293:501-502.

INTRODUCTION

One-third of the world’s population is estimatedby the World Health Organization to be

infected with Mycobacterium tuberculosis (M.tuberculosis). Efforts to control this disease, itstransmission, and ultimately its eradication havebeen fought along two fronts in the United States.The first front is to detect and treat symptomaticpeople with infectious tuberculosis. The goal is tocure these individuals and stop further spread of theorganism. While essential, these efforts arefrequently too late to prevent transmission to otherpeople. The second front is to detect the vast poolof asymptomatic people who have latent M.tuberculosis infection (LTBI) and prevent them fromdeveloping infectious tuberculosis (1).

Tests with high sensitivity and specificitycharacteristics for detectingM. tuberculosis infectioncould facilitate tuberculosis control on both fronts.A sensitive test would facilitate screening of peoplewho would benefit from closer evaluation forinfectious disease or treatment to prevent it fromdeveloping. A specific test would avoid treatingpeople at minimal risk of having or developing theinfectious disease. Attempts to develop and evaluatemore accurate tests have been hampered by the lackof a “gold standard” for identifying latent M.tuberculosis infections.

Until recently, the standard and only method forimmunologic diagnosis of M. tuberculosis infectionhas been limited to the tuberculin skin test (TST).The TST, developed in 1889, has been used to detectboth LTBI and active tuberculosis. However, therehave always been concerns about its shortcomings.Because purified protein derivative of tuberculincontains many antigens that are shared with othermycobacteria, the skin test does not reliablydistinguish LTBI from prior immunization withMycobacterium bovis bacilli Calmette- Guérin (BCG)or infection with environmental mycobacteria (2).This is a major problem in most developed countriesbecause a growing proportion of those with LTBI areforeign born persons from high incidence countries,most of whom received BCG vaccination duringchildhood. False-negative results in the setting ofhost immunosuppression has limited also its utility.In addition, cutaneous sensitivity to tuberculindevelops from 2 to 10 weeks after infection and theTST requires two encounters with a health careprofessional which often causes logistical problemsif not inconvenience. Finally, skilled personnel areessential for proper placement and interpretation ofthe test. A test with greater accuracy andconvenience would greatly enhance tuberculosiscontrol efforts (3).

The new blood assays to detect M. tuberculosisinfection are based on the response of antigen-specific memory T-cells releasing interferon-gamma(IFN-γ) in response to previously encounteredmycobacterial antigens. Interferon-gamma releaseassays (IGRAs) measure the cellular immuneresponses to M. tuberculosis-specific antigens,including early-secreted antigenic target 6 (ESAT-6)and culture filtrate protein 10 (CFP-10), antigensencoded in the region of difference (RD1) of the M.

Interferon-γ Release Assay forDetection of Tuberculosis Infection –Chapter updated in 2011

Alfred Lardizabal, MDDr. Lardizabal is AssociateProfessor of Medicine at NJMedical School-University ofMedicine and Dentistry ofNew Jersey and GlobalTuberculosis Institute inNewark, New Jersey.

tuberculosis genome. These proteins are absent fromall strains of M. bovis BCG and the vast majority ofnon-tuberculous mycobacteria (with the exceptionof M. kansasii, M. szulgai, and M. marinum) butpresent in isolates ofM. tuberculosis. In comparison,the TST uses the mixed, nonspecific PPD, a culturefiltrate of tubercle bacilli containing over 200antigens, which results in its low specificity.

Two IGRA systems using RD1-encoded antigensare currently commercially available for TBdetection. One system includes QuantiFERON®-TBGold and its variant QuantiFERON®-TB Gold In-Tube (uses tubes pre-filled with antigens) (Cellestis,Victoria, Australia), which uses whole bloodspecimens, with an unknown number of leukocytes,to measure IFN-γ released by antigen-activated Tlymphocytes. The other system is the T-SPOT®.TB(Oxford Immunotec, Oxford, England). It uses theELISPOTmethod, wherein the number of peripheralblood mononuclear cells (PBMC) in the assay isquantified, in order to measure IFN-γ-secreting Tcell counts ("spots") on stimulation by M.tuberculosis-specific antigens in microplate wells.The readout of the two tests is different:QuantiFERON®-TB Gold (QFT-G) andQuantiFERON®-TB Gold In-Tube (QFT-GIT)measures the level of IFN-γ in the supernatant of thestimulated whole blood sample using enzyme-linkedimmunosorbent assay (ELISA), and T-SPOT®.TBenumerates individual T-cells producing IFN-γ afterantigenic stimulation.

These new blood tests have an internal positivecontrol, i.e., a sample well stimulated with a potentnon-specific stimulator of IFN-γ production by T-cells. This controls the results of the test fortechnical errors, such as failure to add viable,functioning cells to the well. The failure of thepositive control in the tests provides informationthat the test’s results cannot be reliably interpretedsince it may reflect an underlying in vivoimmunosuppression, negatively affecting T-cellfunction in the in vitro stimulation.

TEST PERFORMANCE

In order to establish the diagnostic accuracy forLTBI of any test is a major challenge because there isno available gold standard. As an alternative, somerational approaches based on the epidemiology ofTB have been applied. The knowledge that airbornetransmission of TB is promoted by close andprolonged contact with an infectious case has beenused with the preposition that if a test is a goodmarker of LTBI, it should correlate closely with thelevel of exposure. Several studies conductedcomparing QFT and TST used in the setting of acontact investigation have results that show thesetests to be moderately concordant (4, 5, 6). Incomparing the IFN-γ assay to the TST in personswith varying risk for MTB infection, Mazurek, et al.(5) showed that the assay was comparable to theTST in its ability to detect LTBI, with an overallagreement of 83%. It was less affected by BCGvaccination, discriminated responses due to NTM,and avoided the variability and subjectivityassociated with placing and reading the TST. Thistest was also used to detect recent infection amongcontacts in a TB outbreak at a Danish high school.Since a majority of contacts were BCG-unvaccinateddirect comparison between the TST and QFT couldbe performed. Analysis revealed an excellentagreement between the two tests was found (94%,kappa value 0.866) and that the blood test was notinfluenced by the vaccination status of the subjectstested (4). Ewer and colleagues investigated a schooloutbreak that resulted from one infectious indexcase using the ELISPOT assay and the Heaf test (7).The overall agreement between the two tests was89%. The ELISPOT assay showed no significantrelation to BCG status. By contrast, BCG-vaccinatedchildren were more likely to have higher Heaf gradesthan unvaccinated children. An isolated positiveELISPOT was associated with exposure, whereas anisolated positive TST result was not. Several studiescompared TST and IGRAs with respect to theircorrelation with exposure to M. tuberculosis (4, 7, 8,9, 10). In these studies the RD1-based assays

showed stronger positive correlation with increasingintensity of exposure compared to the TST.

The sensitivity of IGRAs have been estimatedusing cases of active tuberculosis confirmed bycultures and often excluded HIV-infectedindividuals. Studies that estimated the specificity ofIGRAs were carried out in low incidence countrieswith some patients exposed to BCG vaccination andother not. Pai and colleagues (11) in a recent meta-analysis estimated the pooled sensitivity of the QFTstudies to be 76% and 90% for T-SPOT®.TB. Thepooled specificity for all QFT studies was 98%, (99%for QFT among non-BCG vaccinated populationsand 96% for BCG vaccinated populations). Thepooled specificity of T-SPOT®.TB was 93% (almostall studies included BCG vaccinated participants).For the TST, pooled sensitivity estimate was 77%,specificity in non-BCG vaccinated was 97% but lowand highly heterogeneous among BCG vaccinatedparticipants. From this substantial body of literature,it can be concluded that IGRAs, especially QFT-Gand QFT-GIT, have excellent specificity that isunaffected by BCG vaccination. TST has a highspecificity among non-BCG vaccinated individuals.The sensitivity of IGRAs and TST is not consistentacross populations but the T-SPOT®.TB appears tobe more sensitive than QFT or TST. Similar findingwere observed by Diel and colleagues in a meta-analysis they performed in which the TST had apooled sensitivity of 70%, QFT-IT was 81%, and T-SPOT®.TB was 88%. Specificity of QFT-IT was 99%against 86% for the T-SPOT®.TB (12). Data on high-risk populations, such as immunocompromised andyoung children, remain limited and it has beenshown that indeterminate results for the IGRAs tendto increase in these groups. In addition, one recentstudy which evaluated close contacts using both theIGRA (QFT-IT) and the TST suggested that theIGRA appeared to be a more accurate indicator ofthe presence of LTBI. It also provided some insightinto its predictive value for the development ofactive TB since 14.6% of those with a positive QFT-IT progressed to active TB compared to only 2.3 %among those with a positive TST (13). Questionssuch as the prognostic ability of these tests to

accurately identify individuals with LTBI who are athighest risk for progressing to active TB and thesignificance of conversions and reversions of thesetests over time still need to be clarified.

From the approval by the US Food and DrugAdministration (FDA) of QFT-G in May 2005, theUS Centers for Disease Control and Preventionrecommend that "QFT-G may be used in allcircumstances in which the TST is currently used,including contact investigations, evaluation of recentimmigrants, and sequential testing" with warningsand limitations (14). Updated CDC guidelines forthe use of interferon gamma release assays werepublished in 2010 and states that either the TST andIGRAs (QFT-G, QFT-GIT, T-Spot) may be used asaids in diagnosing M. tuberculosis infection (15).IGRAs are preferred for testing persons who havereceived BCG (as a vaccine or for cancer therapy) orfor testing groups that historically have low rates ofreturning to have TSTs read. The TST is preferredfor testing children aged <5 years. An IGRA or a TSTmay be used without preference for testing recentcontacts to persons with infectious pulmonary TBwith considerations for follow-up testing. An IGRAor a TST may also be used without preference forperiodic screening of persons who might haveoccupational exposure to M. tuberculosis withconsiderations for conversions and reversions.Currently, an IGRA conversion is defined as achange from negative to positive within 2 yearswithout any consideration of the magnitude of thechange in TB response (as opposed to the morestringent 10 mm change required for the TST). Amore stringent criteria for conversion using IGRAs isyet to be defined. Substantial progress has beenmade in documenting the utility of IGRAs butfurther studies and research determining the valueand limitations of IGRAs in situations important tomedical care and TB control is needed.

References:

(1) CDC. Targeted tuberculin testing and treatmentof latent tuberculosis infection. MMWR 2000;49:1-51.

(2) American Thoracic Society. Diagnostic standardsand classification of tuberculosis. Am Rev RespirDis 1990; 142:725-735.

(3) Institute of Medicine. Ending Neglect: theelimination of tuberculosis in the United States.Washington, DC: National Academy Press; 2000.

(4) Brock I, Weldingh K, Lillebaek T, Follmann F,Andersen P. Comparison of a New Specific BloodTest and the Skin Test in Tuberculosis Contacts.Am J Respir Crit Care Med 2004; 170:65-69.

(5) Mazurek GH, LoBue PA, Daley CL, Bernardo J,Lardizabal AA, Bishai WR, Iademarco MF, RothelJS. Comparison of a whole-blood interferon-γassay with tuberculin skin testing for detectinglatent Mycobacterium tuberculosis infection.JAMA 2001; 286:1740-1747.

(6) Streeton JA, Desem N, Jones SL. Sensitivity andspecificity of a gamma interferon blood testtuberculosis infection. Int J Tuberc Lung Dis 1998;2:443-50.

(7) Ewer K, Deeks J, Alvarez L, Bryant G, Waller S,Andersen P, Monk P, Lalvani A. Comparison of T-cell based assay with Tuberculin Skin Testing forthe Diagnosis of Mycobacterial tuberculosisinfection in a School Outbreak. Lancet 2003;361:1168-73.

(8) Hill PC, Brooks RH, Fox A, Fielding K, JeffriesDJ, Jackson-Sillah D, Lugos MD, Owiafe PK,Donkor SA, Hammond AS, Out JK, Corrah T,Adegbola RA, McAdam KP. Large ScaleEvaluation of Enzyme-Linked ImmunospotAssay and Skin Test for Diagnosis ofMycobacterium tuberculosis Infection Against aGradient of Exposure in The Gambia. Clin InfectDis 2004.38:966-73

(9) Lalvani A, Pathan AA, Durkan H, Wilkinson KA,Whelan A, Deeks JJ, ReeceWHH, Latif M, PasvolG, Hill AV. Enhanced Contact Tracing andSpatial Tracking of Mycobacterium tuberculosisInfection by Ennumeration of Antigen SpecificT-cells. Lancet 2001; 357:2017-21.

(10) Richeldi L, Ewer K, Losi M, Bergamini BM,Roversi P, Deeks J, Fabbri LM, Lalvani A. T-cellbased Tracking of Multidrug ResistantTuberculosis Infection After Brief Exposure. AmJ Respir Crit Care Med 2004; 170:288-95.

(11) Pai M, Zwerling A, Menzies D. SystematicReview: T-Cell-Based Assays for the Diagnosisof Latent Tuberculosis Infection: An Update.Ann Intern Med 2008;149:1-8.

(12) Diel R, Loddenkemper R, Nienhaus A.Evidence-based Comparison of CommercialInterferon-γ Release Assays for Detecting ActiveTB. Chest 2010; DOI:10.1378

(13) Diel R, Loddenkemper R, Meywald-Walter K,Niemann S, Nienhaus A. Predictive Value of aWhole Blood IFN- Assay for the Developmentof Active Tuberculosis Disease after RecentInfection withMycobacterium tuberculosis. Am JRespir Crit Care Med 2008; 177:1164-1170.

(14) CDC. Guidelines for using the QuantiFERON-TB Gold test for Detecting Mycobacteriumtuberculosis infection, United States. MMWR2005; 54(RR15):49-55.

(15) CDC. Updated guidelines for using InterferonGamma Release Assays to detectMycobacterium tuberculosis infection – UnitedStates, 2010. MMWR 2010; 59(RR5):1-25.

Karen Galanowsky R.N.,

M.P.H.

Ms. Galanowsky is a Nurse

Consultant for the New Jersey

State Department of Health and

Senior Services, Trenton, New

Jersey, TB Program

The nurse plays a significant role in tuberculosis (TB)

programs, ensuring the control of TB through

targeted tuberculin skin testing (TST) and treatment of

latent TB infection. The nurse’s role intersects all aspects

of public health practice by focusing on individuals,

communities and health care systems.

The nurse integrates the core functions of public

health—assessment, policy development and

assurance—into the interventions and strategies used to

identify individuals at high risk for TB who would benefit

from treatment for latent TB infection if it were detected.

The nursing process is utilized to intervene at all levels—

individual, community or health care system. It includes

assessment, problem identification, planning,

implementation and evaluation.

KNOWLEDGE AND COMPETENCIES

The registered nurse (RN) who works with individuals

at risk for latent TB infection should obtain up-to-date

knowledge on:

� The diagnosis and treatment of latent TB infection,

and the differentiation of TB disease

� The community, including demographic

characteristics, ethnic populations and resources

� How latent TB disease and treatment is perceived

by culturally diverse populations

� Epidemiology TB in the community

THE ROLE OF THE NURSE INDIAGNOSING LATENT TB INFECTION


The registered nurse also should be familiar with:

� Health care providers within the community,

including the populations served, and the strengths

and weaknesses of each

� Barriers to health care—specifically tuberculin skin

testing, reading and interpretation, medical

evaluation and completion of treatment

� Principles of TB control

� Regulations and legal mandates

� Internal and external standards

In addition, the nurse should attain and maintain

proficiency in:

� Administering, reading and interpreting the

tuberculin skin test

� Providing culturally sensitive and age-appropriate

TB education for both individuals and

communities

� Assessing the individual and the community

� Providing interventions and strategies to ensure

that high-risk individuals and groups are

tuberculin skin tested, and that they have access to

medical evaluation and appropriate treatment

� Collaboration

� Networking

� The political process

� Policy development

� Evaluation

INTERVENTIONS

Interventions are actions the nurse takes on behalf of

individuals, families and communities. Particular

interventions are recommended based on the nurse’s

clinical judgment using theoretical, practical and scientific

knowledge. Potential or desired outcomes should be

identified and related to the interventions, which may be

implemented at all levels (individual, community and

systems-focused) and at all sites for TB control.


Interventions fall under a number of categories:

� Surveillance

� Screening

� Patient assessment

� Education (patient, health care provider,

community)

� Counseling

� Advocacy

� Referral and follow-up

� Coalition building

� Collaboration

� Community mobilization

� Consultation

Individual-focused interventions

Interventions that focus on individuals create changes

in the knowledge, attitudes, beliefs, skills, practices and

behaviors of individuals, families and groups. These are

person-to-person interventions specifically for persons

belonging to a population at risk of TB.

The RN generally provides the following individually

focused interventions: screening, patient assessment,

follow-up, monitoring, education, counseling or

consultation. Nurses also conduct contact investigations

and do TST of identified close contacts. This process

involves a series of decisions, frequently made solely by

the nurse. Two examples illustrate the RN as the decision

maker: (1) When a contact investigation is required in a

congregate setting, such as a school, the decision as to

who should be tested, using the concentric circle

principle, is often made by the nurse; and (2) If the

concentric circle needs to be expanded due to an

unexpected number of positive results, the nurse decides

which groups or individuals should be included.

Sometimes the very effort to find and TST identified

contacts takes initiative and ingenuity on the nurse’s part.

The nurse also has to identify other barriers to patients’

being screened and treated for latent TB infection.

With respect to the physician, the RN’s role is typically

one of patient advocacy and educator. By using assertive

diplomacy, the RN is frequently required to provide the

private physician with the most recent literature

regarding the reading and interpretation of the TST and

the recommended treatment for latent TB infection. In

addition, the nurse often educates the patient about the

importance of keeping doctor or clinic appointments so

that treatment may not be interrupted, and explains

potential side effects of medication. The nurse also

explains TB infection and the difference between TB

infection and disease. Frequently, it is the nurse who

monitors the patient for side effects of medication, and

tracks the patient to ensure completion of treatment.

Problems encountered

A number of problems have arisen in connection with

individual-focused interventions. Private physicians or

patients themselves may be apathetic about the

importance of TST. Patients may fail to return to have their

TST read. All too often, patients and even private

physicians may believe that BCG is the cause of the

positive TST result. Physicians may not prescribe any

treatment, while patients may not adhere to treatment even

if it is provided.

The health care provider may fail to assess for TB

symptoms when the TST is negative, even though the

patient may belong to a population at heightened risk of

TB. The patient may have personal beliefs about health

or attitudes about TB that interfere with care. The patient

may not understand the difference between TB infection

and TB disease. He/she may fear side effects or adverse

reactions of TB treatment.

Other problems arise as well. Co-existing medical or

psychiatric diagnoses may impair the patient’s ability to

give accurate information. For many persons at risk of

TB, substance abuse, homelessness or residential

instability may inhibit efforts to intervene. The patient

may be faced with competing or conflicting demands or

may lack the funds to pay for medical care.

Case presentation: “The Cliff Dweller”

The patient is a 39-year-old female, born in South

America, who immigrated to the United States twenty years

ago. She was first diagnosed with pulmonary TB during a

hospitalization for injuries she sustained from an assault by

a boyfriend. At the time her clinical picture revealed:

� History of positive TST (date unknown)

� Symptoms of cough, hemoptysis and weight loss

� Chest x-ray: Cavity right apex, diffuse infiltrates

left upper lobe

� Sputum smear negative; culture M. tuberculosis

The patient gave an address of an apartment in a local

municipality. She named three contacts at the same

address, and said she had one daughter who lived nearby.


The RN tried to locate and interview the named contacts,

but found that the address the woman provided was

fictitious. When the RN revisited the hospital to talk

with the patient, the patient had signed herself out

against medical advice.

The RN then went to locate the patient in the

community. Because all leads resulted in dead ends, the

RN questioned the police department as to the patient’s

whereabouts. The police had multiple dealings with the

patient, who came in as the victim of physical and sexual

assault by her boyfriend. They provided four different

addresses including two liquor stores and “the cliffs.”

The police were not aware that anyone was living on the

cliffs, a steep and wooded embankment sloping down to

a major highway.

The patient became lost to follow-up for six months

until she was again assaulted and required

hospitalization. The RN interviewed the patient

immediately. At this time, the patient was considered

highly infectious. The patient stated that she lived in

tents on the cliffs. She said she slept in a different tent

each night in exchange for sexual favors. During the day

she went to the streets where she “hung out.”

Occasionally she went to her daughters to shower.

The RN facilitated the use of legal intervention to

constrain the patient in the hospital for treatment until

smear negative results were obtained. During this time,

the RN visited the patient daily, brought food the patient

requested and assisted the patient to obtain a television in

her room. The RN’s goal was to develop a working

relationship with both the patient and her boyfriend over

the next several weeks. The boyfriend told the RN that

Saturday mornings would be a good time to find all the

people who stay on the cliffs. A team, including a field

supervisor, field worker, nurse and translator, went to the

cliffs on a Saturday morning. They brought hot coffee,

rolls and doughnuts as “peace-bearing gifts.” They also

brought TST material.

The team found and tuberculin tested 13 cliff dwellers.

The team arranged to pick up the group of contacts by

van and bring them to the clinic on Monday morning for

a TST reading and further medical evaluation. The team

promised incentives for adherence. Five of the patients’

relatives were also located and tested over the following

two weeks with the assistance of the patient’s boyfriend.

Interventions utilized

Contacts came to the clinic drunk, disorderly and

extremely dirty. Clinic staff provided transportation and

remained nonjudgmental and kind. Incentives such as

food, gift certificates for food, and clothing were given to

all the contacts at the site where they dwelled. Social

service intervention also was provided to assist them in

finding better housing.

Education was provided to the contacts verbally, slowly

and over time. Most of them could not read or write.

Team members discussed and allayed the contacts’ fear of

deportation and mistrust of the government; most of

them were undocumented immigrants. Behavioral

contracts were developed verbally with the contacts to

ensure their adherence to directly observed treatment for

latent TB infection.

Nursing and field staff had to locate the contacts by

traversing the cliffs through mud, trees and debris. The

TB control RN worked with the police and mayor to avert

political pressure to have the cliff dwellers removed

immediately. The RN also worked with the public health

officer and the press to prevent panicked coverage by

newspapers and television.

Community-focused interventions

The ultimate goal of community-focused nursing

interventions is to prevent the development of TB disease

in persons with latent TB infection in the community.

The nurse accomplishes this goal by:

� Networking

� Coalition building

� Teaching other nurses and health care workers

� Community assessment

� Collaboration

� Community education

� Consultation

� Evaluation

With an understanding of the incidence and

prevalence of TB disease, and the socio-demographic

characteristics of a community, the nurse can identify

those individuals and groups who are likely to be

infected with TB and benefit from treatment. The nurse

can establish partnerships with community health

centers, outpatient clinics, drug treatment centers and

homeless shelters that provide services to individuals at

high risk for TB infection and disease. The nurse also can

provide TB education to the community and health care

providers to increase awareness of TB.

Community-focused interventions require that nurses

are proficient in administering and reading the TST.


Assurance that there are appropriate resources to

medically evaluate and treat those individuals who are

identified as TST-positive is also important.

Community-focused interventions also include the

evaluation of those nurses administering and interpreting

the TSTs within the community; training should to be

provided to assure and maintain proficiency in this

activity. Evaluation also focuses on programmatic

outcomes of treatment.

A community-based TB prevention model

Another step the nurse can take in identifying those at

risk, diagnosing latent TB infection and connecting

patients to appropriate treatment and care is to

implement a community-based model for the delivery of

TB preventive services that accommodate the

community’s characteristics and needs.

An excellent example of a community-based model of

TB preventive services was developed by the TB program

at the Boston Public Health Department, Boston, Mass.

This model entailed collaboration between the health

centers and the Boston Public Health Department’s TB

program. A nurse coordinated the program. The health

centers identified “core” provider teams who became the

TB expert resources for individual health center

programs, and the health department’s TB program

educated and trained the core provider team for each

health center.

Policies for TB screening of patients and staff were

reviewed and revised collaboratively, and were based on

the needs of the community health centers. Persons with

a positive TST were referred to the Boston Public TB

Clinic for an initial medical evaluation and treatment.

Those who were placed on treatment for latent TB

infection were monitored monthly and given a new

supply of TB medication at the community health center.

The health department TB program delivered the

monthly supply of TB medications to each health center

and evaluated patient adherence, monitoring and

documentation. Data were collected and analyzed by the

TB program and shared with the health centers to help

identify obstacles to the implementation of this plan and to

document accomplishments in meeting specific objectives.

Problems encountered

A number of problems regularly arise in working with

the community. Poor documentation of TST results is

common (recording “negative” or “positive” rather than

the millimeter reading.) Measuring redness and/or

swelling rather than induration is another frequently

found problem. Still other common problems include:

� Measuring with a standard ruler—or guessing

� Making a determination based on the appearance

of redness without palpating the arm

� Filling the syringe with the PPD hours before

administering the TST

� Inadequate storage of PPD, not refrigerating it or

keeping it out in the sunlight all day

� Delay in referring an individual with a positive TST

result for medical evaluation without assessing

symptoms of TB disease

� No treatment for LTBI recommended

� “Un-targeted” testing (e.g., TST of all patients

admitted to a psychiatric inpatient unit for no

apparent reason)

� Inadequate community resources, or barriers

inherent in the managed care system, for adequate,

timely follow-up of TST-positive individuals

� Apathy regarding positive TST results on the part

of individuals and health care providers

� Physician’s lack of knowledge regarding current

treatment recommendations

� Hysteria about TB and lack of knowledge about the

difference between TB infection and disease

Some of the above problems can be alleviated with the

use of QuantiFERON®-TB Gold In-Tube. QuantiFERON®-

TB Gold In-Tube has the advantage of minimizing error due

to interpretation on the part of the reader. The patient also

does not have to be followed up for a second visit for

interpretation of the result.

CONCLUSION

Nurses play a vital role in TB programs, ensuring that

targeted tuberculin skin testing is provided to high-risk

individuals within a community. Without skillful

nursing interventions, many individuals would not be

tuberculin tested and those who are positive would not

be medically evaluated, diagnosed and treated. Nurses

are faced with multiple challenges from the individual,

community and other health care providers. Confronted

with these challenges, nurses prevail by using the nursing

process and principles of public health practice to

achieve significant outcomes.


antigen—any substance that is capable, under

appropriate conditions, of inducing the formation of

antibodies and the sensitization of lymphocytes and of

reacting specifically in some detectable manner with the

antibodies or lymphocytes so induced.

Battey antigen (PPD-B)—a PPD culture filtrate made

from Mycobacterium intracellulare that may produce

tuberculosis-like disease in humans and may induce a

positive response to the Mantoux test.

conversion, Mantoux conversions—a term denoting

the change from a tuberculin-negative to a tuberculin-

positive state, as determined by the Mantoux test.

cross-sensitization—sensitization to a substance

induced by exposure to another substance having cross-

reacting antigens.

gamma interferon release assay—measures the cell

mediated response in infected individuals through the

levels of interferon gamma released.

Mantoux test—an intracutaneous injection of 0.1 ml of

tuberculin and assessment of any reaction, in size of

induration (in mm).

PPD-B—see Battey antigen.

PPD-S—a large batch of tuberculin (No. 49608) that

was prepared by Seibert in 1939 and has been adopted as

the international and U.S. reference standard tuberculin.

purified protein derivative (PPD)—a sterile, soluble,

partially purified product derived from a tubercle bacillus

culture. It is used as a dermal reactivity indicator in the

diagnosis of tuberculosis.

5 TU (tuberculin units)—a dose of tuberculin that is

biologically equivalent to that contained in 5 TU PPD-S.

tuberculin—a sterile liquid containing the growth

products of, or specific substances extracted from, the

tubercle bacillus.

Tween®—trademark for a sorbitan polyoxyalkalene

derivative; used as an emulsifier and detergent.

Tween 80®—trademark for polysorbate 80.

GLOSSARY


1. One of the most essential components of the U.S. TB elimination strategy is:a. BCG Vaccinationb. Targeted tuberculin testingc. Isoniazid chemoprophylaxisd. Targeted tuberculin testing and treatment for latent TB infectione. Annual chest x-rays

2. Major problems in addressing TB control in low incidence settings include all except:a. Drug resistanceb. Microepidemicsc. Lack of professional expertised. Unfamiliarity with treatment of latent TB infectione. Lack of sanitarium beds

3. Tuberculin skin testing is useful for all but:a. Contact tracingb. Screening of risk groups for TBc. Anergy testingd. Measuring annual risk of infection and program supporte. Skin testing of symptomatic patients

4. Tuberculins are:a. Dead tubercle bacillib. Serum from TB infected patientsc. Mixtures of culture filtrate components of sterilized cultures of tubercle bacillid. Unstable antigens

5. The recent study that assessed variability of commercially available tuberculins revealed all but:a. Specificity of Aplisol® and Tubersol® were equally high and similar to PPD-Sb. Testing with Tubersol® produced slightly smaller reactions and Aplisol® slightly larger reactions than PPD-S,

but these differences did not affect TST interpretationc. Both Aplisol® and Tubersol® correctly classified comparable number of persons not infected with TBd. Either commercial product may be used with confidence for tuberculin skin testinge. The products can easily be interchanged in serial tests

6. The overriding rule for administering and reading tuberculin tests is:a. Everybody should have an annual tuberculin testb. Test only those in when therapy for latent TB infection is indicated – the high risk reactorc. Test only children under 5d. Test only individuals under 35 years of agee. None of the above

7. Tuberculin conversion is defined as:a. An increase in reaction size of more than 2 mmb. An increase in reaction size of more than 4 mmc. An increase in reaction size of more than 6 mmd. Any new positive reaction in an individual with a history of a negative reaction

8. The Booster Phenomenon is all except:a. Is often seen in elderly personsb. Defined as an increase in tuberculin skin reactions following repeat tuberculin testing unrelated to new

mycobacterial infectionc. Occurs when a waned TST reading is stimulated by a tuberculin testd. Denotes immunosuppressione. Can often be ruled out by repeating a negative test in a week

9. Which statement is true?a. Accurate assessment of non-tuberculous mycobacterial infection can be made by skin test reaction sizeb. Antigens made from non-tuberculous mycobacteria are commercially available c. Non-tuberculous mycobacteria can be prevented by targeted tuberculin testing and treatmentd. Cross reactions to non-tuberculous mycobacteria may be responsible for small tuberculin reactions

10. All of the following are true about the QuantiFERON®-TB In-Tube test except for: a. Requires 1 patient visit to a clinician for administration and interpretationb. Is an in vivo testc. Results are possible in one dayd. There is no boostinge. Requires phlebotomy

ACTIVITY POST-TEST (record your answers on the registration form)


UNIVERSITY OF MEDICINE AND DENTISTRY OF NEW JERSEY

CENTER FOR CONTINUING AND OUTREACH EDUCATION

GUIDELINES FOR THE DIAGNOSIS OF TUBERCULOSIS INFECTION IN THE 21ST CENTURY

Registration FormIn order to obtain continuing education credit, participants are is required to

1. Read the learning objectives, review the publication and complete the post-test.

2. Complete this registration form, record your test answers in the box below, and complete the activity evaluation

form on the reverse side.

3. Send the registration and evaluation forms to

UMDNJ-Center for Continuing and Outreach Education

via mail: P.O. Box 1709, Newark, NJ 07101-1709

via fax: (973) 972-7128

4. Retain a copy of your test answers. Your answer sheet will be graded and if a passing score of 70% or more is

achieved, a continuing education credit letter and test answer key will be mailed to you within four weeks.

Individuals who fail to attain a passing score will be notified and offered to complete the activity again.

PLEASE PRINT CLEARLY

First Name___________________________ M.I. _______ Last Name_________________________ Degree(s) ______

Last 4 digits of Social Security # ________________________________________________________________________(for CE office records)

Daytime Phone ___________________________________ Evening Phone _____________________ ______________

Fax_____________________________________________ E-mail ____________________________

Preferred Mailing Address � Home � Business

__________________________________________________________________________________________________

__________________________________________________________________________________________________

City ____________________________________________ State _________________ Zip Code __________________

Affiliation _______________________________________ Specialty__________________________________________

Indicate the type of continuing education credit you wish to obtain as a result of your participation in this activity.

� Physicians: AMA PRA Category 1 Credit(s)™ Credit Letter: Credits Claimed: __________

� Nurses: Nursing Contact Hours (ANCC): Hours Awarded: 3.5

� Other: Certificate of Participation: Credits Claimed: __________

One (1) credit for each hour of participation; not to exceed 3.5 credits.

I attest that I have completed this activity as designed. I will report the number of credits claimed above during my filing

of continuing education credit with professional organizations, licensing boards, or other agencies.

Signature Date

Claiming credit for this activity is available through March 31, 2010.

UMDNJ-Center for Continuing and Outreach Education, P.O. Box 1709, Newark, NJ 07101-1709

Phone: 973-972-4267 or 1-800-227-4852

CE Activity Code: 10MN02

POST-TESTCircle the best answer for each question on the facing page.

1. A B C D E 6. A B C D E2. A B C D E 7. A B C D3. A B C D E 8. A B C D E4. A B C D 9. A B C D5. A B C D E 10. A B C D E


UNIVERSITY OF MEDICINE AND DENTISTRY OF NEW JERSEY

CENTER FOR CONTINUING AND OUTREACH EDUCATION

GUIDELINES FOR THE DIAGNOSIS OF TUBERCULOSIS INFECTION IN THE 21ST CENTURY

ACTIVITY EVALUATION FORMThe planning and execution of useful and educationally sound continuing education (CE) activities are guided in large part

by input from learners. To assist us in evaluating the effectiveness of this activity and to make recommendations for future

educational offerings, please take a few moments to complete this evaluation form. Your response will help ensure that

future programs are informative and meet the educational needs of all participants. Please note: CE credit letters will only

be issued only receipt of a completed evaluation form. Thank you for your cooperation!

Please indicate your profession:

� Physician � Nurse � Physician’s Assistant � Other _____________________________

PROGRAM OBJECTIVES Strongly Strongly Agree Disagree

Having completed this activity, are you better able to:

Describe the role of tuberculin testing in low prevalence countries. 5 4 3 2 1

Review how tuberculins are developed, manufactured and validated. 5 4 3 2 1

Recognize minor disparities in commercially available tuberculins and the necessity of

serial testing with the same antigen. 5 4 3 2 1

Explain the protocol for administering and reading tuberculin skin tests. 5 4 3 2 1

Correctly interpret repeated tuberculin skin tests. 5 4 3 2 1

Examine the role of tuberculin reactions produced by cross reactions

with non-tuberculous mycobacteria. 5 4 3 2 1

Differentiate the use of interferon-γ release assays test when compared to tuberculin skin testing. 5 4 3 2 1

Discuss the role of the nurse in the diagnosis of latent TB infection. 5 4 3 2 1

OVERALL EVALUATION Strongly Strongly Agree Disagree

The information presented increased my awareness/understanding of the subject. 5 4 3 2 1

The information presented will influence how I practice. 5 4 3 2 1

The information presented will help me improve patient care. 5 4 3 2 1

The faculty demonstrated current knowledge of the subject. 5 4 3 2 1

The information presented was educationally sound and scientifically balanced. 5 4 3 2 1

The information presented avoided commercial bias or influence. 5 4 3 2 1

Overall, this activity met my expectations. 5 4 3 2 1

I would recommend this activity to my colleagues. 5 4 3 2 1

If you anticipate changing one or more aspects of your practice as a result of your participation in this activity, please

provide us with a brief description of how you plan to do so.

Please provide any additional comments pertaining to this activity (positive and negative) and suggestions for improvement.

Please list any topics that you would like to be addressed in future educational activities:

CE Activity Code: 10MN02



PO Box 1709 • 225 Warren Street • Newark, NJ 07101-1709

CME Certified Monograph2nd Edition

Documents

GUIDELINES FOR THE DIAGNOSIS OF LATENT TUBERCULOSIS INFECTIONglobaltb.njms.rutgers.edu/downloads/products/guideltbi.pdf · guidelines for the diagnosis of latent tuberculosis infection